Eastman Kodak Company v. Agfa-Gevaert N.V

Full title: EASTMAN KODAK COMPANY, Plaintiff, v. AGFA-GEVAERT N.V. and AGFA CORP.…

Court: United States District Court, W.D. New York

Date published: Apr 22, 2008

560 F. Supp. 2d 227 (W.D.N.Y. 2008)

Opinion

No. 02-CV-6564T.

April 22, 2008

Kristen Mollnow Walsh, Michael F. Orman, Richard D. Rochford, Nixon Peabody LLP, Rochester, NY, Maia H. Harris, Nixon Peabody LLP, Boston, MA, Robert P. Fletcher, Nixon Peabody LLP, Washington, DC, for Plaintiff.

H. Michael Hartmann, Jeffrey B. Burgan, Leydig, Voit Mayer, Ltd., Chicago, IL, Stephen G. Schwarz, Faraci Lange LLP, Rochester, NY, for Defendants.

DECISION and ORDER

MICHAEL TELESCA, Senior District Judge

By Order dated September 26, 2006, upon consent of the parties, I appointed Special Master Joseph W. Berenato, III, as a Special Trial Master in this matter for the purpose of conducting a bench trial and issuing a Report and Recommendation to the court as to how the issues of liability in this case should be resolved. From May 3 to May 11, 2007, Special Master Berenato conducted the bench trial of this matter, and on September 7, 2007, he issued a thorough, well-reasoned 122 page Report and Recommendation recommending that the court find: (1) that the defendants' products Cronex 10TL and Orthovision G infringe plaintiff's U.S. Patents Nos. 4,425,425, 4,425,426, and 4,439,520; (2) that the remaining accused products do not infringe plaintiff's patents; (3) that the asserted claims of the plaintiff's patents are valid; (4) that the asserted claims of the plaintiff's patents are enforceable; (5) that the defendants' infringement of plaintiff's patents was not willful, and (6) that plaintiff failed to provide defendant with notice of infringement with respect to Cronex 10TL prior to expiration of the patents in suit, and therefore, plaintiff is not entitled to damages for the infringement of those patents with respect to Cronex 10TL.

On September 27, 2007, and October 2, 2007, respectively, Special Master Berena-to issued two additional Reports with respect to evidentiary issues and trial matters. Thereafter, in accordance with the court's September 26, 2006 Order, the parties filed objections to the Special Master's Reports.

Following a review of the record de novo, for the reasons set forth in Special Master Berenato's September 7, 2007 Report and Recommendation, I adopt that Report and Recommendation in its entirety without modification. As an initial matter, I find no legal error in any aspect of the Special Master's Report and Recommendation. The authority relied on by the Special Master is appropriate and controlling, and I find no error in the Special Master's legal analysis or conclusions.

With respect to the Special Master's development and discussion of the factual record in this matter, his analysis of the facts is so comprehensive and so complete, additional discussion of the factual issues would serve no fruitful purpose. In short, Special Master Berenato provided the court and the parties with an extensive and detailed explanation as to how and why he arrived at his recommendations. Because I agree in whole with his Report and Recommendation, I adopt his September 7, 2007 Report and Recommendation in its entirety, and make it a Final Order of this Court. With respect to the parties' objections, I find that the issues raised in their objections were thoroughly explained by Special Master Berenato in his Report and Recommendation, and accordingly, I deny those objections without further comment. Finally, for the reasons stated in the Special Master's September 27, 2007 and October 2, 2007 Reports, I adopt those Reports in their entirety.

ALL OF THE ABOVE IS SO ORDERED.

REPORT AND RECOMMENDATION

INTRODUCTION

Pursuant to the Court's Order of September 26, 2006, the undersigned was appointed Special Master to issue a Report and Recommendation on the construction of the claims and, upon agreement of the parties, to issue a Report and Recommendation on the liability issues. The Report and Recommendation for Construction of Group I Patents was filed March 21, 2007. Trial was held May 3-11, 2007. Thereafter, the parties filed post-trial findings of fact and conclusions of law. The parties also submitted a list of exhibits as to which they agreed there were no objections to admissibility. Additionally, Defendants filed a post-trial motion seeking the admission of an exhibit to which Kodak had objected. A Report and Recommendation on that motion and the objected to exhibits is being separately filed.

Having heard the witnesses and considered the parties submissions, the following constitutes my Report and Recommendation regarding the ultimate issues of liability.

Background

As originally filed, this case involved alleged infringement by Defendants Agfa-Gevaert N.V. and Agfa Corporation (collectively "Agfa") of seven patents owned by Plaintiff Eastman Kodak Company ("Kodak"). By stipulation of the parties, the so-called cross-over patents were eliminated from the case and trial proceeded based upon the T-grain patents. The T-grain patents are U.S. Patents Nos. 4,425, 425 ("the `425 patent"), 4,439,520 ("the `520 patent"), and 4,425,426 ("the `426 patent"). The `425 and `520 patents expired November 12, 2001 and the `426 patent expired September 30, 2002. Moreover, the `426 patent was the subject of a patent reexamination, Reexamination Control No. 90/001,255, filed June 8, 1987, by E.I. du Pont de Nemours Company ("DuPont"), predecessor in interest to Defendants. As a result of the reexamination, no claims were canceled or amended although Kodak submitted affidavits and remarks distinguishing certain prior art.

Subsequent to the reexamination proceeding, DuPont spun-off its X-ray film business interests in 1996 as Sterling Diagnostic Imaging, Inc. ("Sterling"). In 1999, Agfa acquired Sterling and assumed Sterling's patent infringement liabilities. Trial Tr. 669 (Verhoeven); TX 99 (Agfa Corp. Depo. Exh. 53 § 2.03). During its existence, Sterling manufactured X-ray film products at a plant in Brevard, North Carolina and sold such products in the United States. TX 823 (Response to Interrogatory No. 1 at 10), 28 (Agfa Corp. Depo. Exh. 48).

Kodak contends that claims 1-3, 5-9, and 19 of the `425 patent are infringed, that claims 1-5, 8-10, 13, 16, 19, 20, and 28 of the `520 patent are infringed, and that claims 1-8 of the `426 patent are infringed. Trial proceeded on the following products: Sterling Ultravision Ci, Sterling Ultravision L, Sterling Ultravision C, Agfa CP-BU, Agfa Ortho HT-G, Sterling Cronex 10T, Sterling Cronex 10TL, Sterling Orthovision L, Sterling Orthovision G, DI Image Plus Green, and Agfa Ortho Opthos H.¹ Kodak asserts both literal infringement and infringement under the doctrine of equivalents. Defendants have denied infringement and have asserted the defenses of invalidity, non-infringement, laches, and unenforceability, and, with respect to Cronex 10TL, lack of notice under 35 U.S.C. § 287. Kodak furthermore contends that the infringement has been willful, which Defendants deny.

1 The Court earlier ruled that Kodak was not entitled to a claim for damages by the Ultravision G product because Kodak's products were not marked and no timely notice of infringement under 35 U.S.C. § 287 was provided to Agfa.

Notice as to Sterling Cronex 10TL

Kodak has accused both Cronex 10T and Cronex 10TL of infringing the `426 patent. Although denying that Cronex 10TL infringes, Defendants argue that they are not liable for any infringement because Kodak failed to mark its products with the patent numbers of the T-grain patents as required by 35 U.S.C. § 287, that they did not receive notice of infringement by Cronex 10TL until after the T-grain patents had expired, and that consequently they cannot be liable for any infringement by that product. Rite-Hite Corp. v. Kelley Co., 56 F.3d 1538, 1549 n. 8 (Fed. Cir. 1995) (en banc), cert. denied, 516 U.S. 867, 116 S.Ct. 184, 133 L.Ed.2d 122 (1995).

35 U.S.C. § 287(a) provides generally that where the manufacturer or seller of a patented article fails to mark the article or its packaging with a notice that the item is patented, the patentee shall not be entitled to damages for infringement absent "proof that the infringer was notified of the infringement and continued to infringe thereafter. . . ." The section further provides that where the patentee has not marked a patented article, damages for infringement are limited to the period following the date on which the infringer received actual notice of infringement. 35 U.S.C. § 287(a).

There is no dispute that Kodak did not mark its X-ray products with a notice that they were patented under one or more of the patents in suit. Kodak has not disputed that for purposes of calculating damages, the damages period commenced when Agfa, or its predecessors in interest, received actual notice of alleged infringement of the patents in suit.

Kodak's original Complaint alleged that Cronex 10T and a number of other Agfa film products infringed Kodak's `426, `425, and/or `520 patents. Although various products were identified as infringing, the original Complaint did not identify Cronex 10TL as infringing. TX 1040 (¶ 16). It was not until October 24, 2004, i.e., after expiration of the T-grain patents, that Kodak supplemented its response to Agfa's Interrogatory No. 5 and specifically accused Cronex 10TL of infringement. On July 13, 2005, Kodak filed an Amended Complaint and therein accused Cronex 10TL and other products of infringement. TX 1058.

In a Decision of July 2, 2004, this Court held that Kodak provided actual notice as to the Cronex 10T product in correspondence to DuPont dated September 7, 1995. The Court further held that the 1995 correspondence did not implicate any other accused product, with the possible exception of Cronex 10TL. The Court found that material disputes of fact regarding similarities between Cronex 10T and Cronex 10TL precluded grant of summary judgment on whether notice as to Cronex 10T was sufficient as to Cronex 10TL as well.

Kodak contends that notice it provided Sterling, Agfa's predecessor in interest, in 1995 regarding infringement by Cronex 10T was sufficient to place Sterling and its assigns on notice that Cronex 10TL infringed the `426 patent. According to Kodak, Cronex 10TL is the latitude version of Cronex 10T and is thus a mere variation of Cronex 10T. Agfa denies that Cronex 10TL is a minor variation of Cronex 10T. Agfa asserts that the September 7, 1995 correspondence did not provide actual notice with respect to Cronex 10TL. According to Agfa, actual notice with respect to Cronex 10TL was not provided until after the T-grain patents had expired, and therefore 35 U.S.C. § 287(a) precludes damages for any infringement by Cronex 10TL.

Cronex 10T is classified as a "gradient" film, whereas Cronex 10TL is classified as a "latitude" film. According to Agfa, it was well understood and widely recognized that gradient films were designed for applications requiring clear views of bone but offering little information for diagnostic purposes about the adjacent soft tissues, such as the lungs or mediastinum. TX 1342 (¶ 15); Trial Tr. at 682-83, 685-88, 691-93 (Hurwitz). Latitude films, on the other hand, were widely accepted as designed to capture images from a broader range of exposure levels than a gradient film, e.g., Cronex 10T film. TX 1060 (¶ 32), 1120; Trial Tr. at 682-83, 685-88, 691-93 (Hurwitz), 841-42 (Hudson). The differences in sensitometric properties between the films led to the universally accepted use of the letter "L" in trade names such as Cronex 10TL to denominate latitude films. TX 1060 (¶¶ 34, 35).

By letter of September 7, 1995, Kodak notified DuPont that Cronex 10T infringed the `425 and `426 patents², and attached a Cronex 10T brochure. TX 150. By letter of October 11, 1995 DuPont denied infringement and suggested a meeting to discuss the matter. TX 431. Kodak thereafter notified DuPont that it was willing to meet to discuss its infringement concerns. The parties met in November 1995 at DuPont's facility in Brevard, North Carolina. Two participants at that meeting, Messrs. Hawley and Waaser, testified that only Cronex 10T was discussed at the Kodak/DuPont meeting and that Kodak never mentioned Cronex 10TL at the meeting or in any other communication with DuPont or DuPont's successor Sterling. Trial Tr. at 514-15, 521-24, 555 (Hawley), 572-73, 576-77, 631 (Waaser).

2 At trial, Kodak asserted that Cronex 10T and Cronex 10TL infringed only the `426 patent.

Q. What technical data of films, other than Cronex 10T, was discussed at the meeting?

A. No technical data was discussed regarding other films.

Q. In fact, you said, Mr. Hawley, you were unaware of any technical data about any film other than Cronex 10T both in September and in November of 1995, correct?

A. Personally that's correct.

Trial Tr. at 555 (Hawley).

Q. Now, either at the meeting or in this letter, did you provide any coding information or batch numbers for any product other than Cronex 10T?

A. No.

Trial Tr. at 584-85 (Waaser).

The parties agree that DuPont explained at the meeting that it was aware of the T-grain patents, that it designed and manufactured the products to avoid infringement, but that start-up issues at DuPont's Brevard facility may have caused some out-of-specification product to be produced. DuPont assured Kodak that it had corrected the start-up issues and provided Kodak with production codes so that Kodak could acquire in the market Cronex 10T specimens, determine that their date of manufacture was after the meeting, and conclude that the product did not infringe the T-grain patents.

Kodak did acquire and test specimens of Cronex 10T after the November 1995 meeting. Kodak acquired and tested a sample of Cronex 10T in May 1996. There was no testimony about Kodak testing of Cronex 10TL for infringement of the T-grain patents as a result of the meeting. Kodak had in June 1995 tested Cronex 10TL and yet the September 1995 letter accused only Cronex 10T of infringement. TX 150, 1246, 1341 (¶ 214).

In an April 9, 2001 letter, i.e., almost six years later, Kodak informed Agfa that it believed that Cronex 10T and ten other Agfa and former Sterling/DuPont products infringed at least one or more of Kodak's `426, `425, and `520 patents. The April 9, 2001 letter identified eleven film products but made no reference to Cronex 10TL as an allegedly infringing product. TX 432, 1281; Trial Tr. at 641 (Verhoeven), 715-16 (Goedeweeck). The letter identified distinctly denominated products within the same product family, e.g., Ultravision Ci, Ultravision L, and Ultravision C, and Orthovision L and Orthovision G. Testimony at trial established that it was industry practice to designate latitude films with the letter "L", so the 2001 letter makes clear that Kodak understood this practice and thus identified specific Ultravision and Orthovision products individually, and not generically by family.

"Actual notice requires the affirmative communication of a specific charge of infringement by a specific accused product or device." Amsted Indus. Inc. v. Buckeye Steel Castings Co., 24 F.3d 178, 187 (Fed. Cir. 1994). Resolution of whether notice is sufficient requires a determination of whether "the recipient is informed of the identity of the patent and the activity that is believed to be an infringement." SRI Int'l v. Advanced Tech. Labs., 127 F.3d 1462, 1470 (Fed. Cir. 1997).

In Gart v. Logitech, Inc., 254 F.3d 1334 (Fed. Cir. 2001), cert. denied, 534 U.S. 1114, 122 S.Ct. 921, 151 L.Ed.2d 886 (2002), cited by both parties, the patentee Gart sent Logitech's counsel a 1995 letter suggesting that Logitech's counsel consider Gart's patent to determine whether a license was needed with respect to TRACKMAN VISTA, a product of Logitech. Gart, 254 F.3d at 1337. In a 1996 letter patentee Gart cited its patent to Logitech as "particularly interesting relative to [Logitech's] trackball product . . . being sold under the trademark TRACKMEN [sic — TRACKMAN] VISTA and TRACKMEN [sic — TRACKMAN] MARBLE." Id. at 1338. In a 1997 letter, Gart stated it was investigating infringement of "Vista and Marble Trackballs [sic]" products. Id. Gart filed a complaint in 1998 for infringement by MOUSEMAN, MOUSEMAN +, TRACKMAN VISTA, TRACKMAN MARBLE, TRACKMAN MARBLE +, and TRACKMAN MARBLE FX. Id.

The Federal Circuit in Gart found that the 1995 and 1996 letters provided actual notice as to TRACKMAN VISTA and TRACKMAN MARBLE, respectively. Gart, 254 F.3d at 1346-47. The Federal Circuit also found that notice was sufficient as to the MARBLE +, and MARBLE FX products as of the 1996 letter, even though the 1996 letter only mentioned the TRACKMAN MARBLE product. Id. at 1347. However, the letters did not serve as actual notice of infringement with respect to the MOUSEMAN and MOUSEMAN + products. Id. An examination of the predecessor and subsequent decisions does not reveal any further details concerning differences between the various products which might shed more light on the Federal Circuit's holding. See Tenneco Automotive Operating Co. v. Visteon Corp., 375 F.Supp.2d 360, 365 (D.Del. 2005) (citing Gart).

The Gart decision has been construed by one district court as follows:

[T]he Court cannot accept Defendants' reading of Gart to foreclose actual notice from ever applying to any infringing product that exhibits anything other than complete and total identity with one cited in the "notice" at issue. Although the Federal Circuit apparently did reject the appellant's argument that certain allegedly similar (but distinctly denominated) devices were encompassed by the notice provided to the patentee, it did not provide an explanation in support of that holding, and the Court will not ascribe to that rejection a meaning not much as alluded to by the Federal Circuit. See 254 F.3d at 1346-47. Indeed, the Court notes that the Federal Circuit's statement that "we hold that the district court erroneously found that neither the 1995 letter nor the 1996 letter effected actual notice under section 287, at least as to the products referenced in those letter[,]" suggests an openness to the possibility that certain products not particularly referenced in the subject communications could be held to fall within the charge of infringement levied in the communications. See id. at 1346 (emphasis added).

Coca — Cola Co. v. Pepsico, Inc., 2004 WL 4910334 n. 17, 2004 U.S. Dist. LEXIS 30375 n. 17 (N.D.Ga. 2004).

The district court in Coca-Cola Co. applied what it termed an "intermediate standard" of actual notice:

[T]he requirement of notice of infringement directed towards a "specific accused product or device" is satisfied where the accused device is merely a modification or evolution of that referenced in the notice of infringement and fails to exhibit a substantial functional change from the device so referenced. Such a standard, in the view of the Court, pays due regard to the rights of accused infringers to have a sufficiently clear understanding of what "specific product" (as opposed to a "family" or "class" of products) is alleged to be infringing, while at the same time avoiding the imposition upon patentees of the burden of constantly evaluating each embodiment of a competitor's product to determine whether a minute "tweak" in design mandates a supplemental notice of infringement.

Coca-Cola Co., 2004 WL 4910334, at *30, 204 U.S. Dist. LEXIS 30375, at 93-94.

Physically, the imaging layers of the Cronex 10T and 10TL were formulated using different silver halide grains. TX 1341 (¶ 211); Trial Tr. at 839-43 (Hudson). From its inception, Cronex 10T had contained a single type — as opposed to a blend — of tabular grains in its imaging layers. Before 1997, Cronex 10T employed a Type 88 grain that had an average grain volume of about 0.24 cubic micron. In 1997, after the DuPont business had been spun off, Sterling began to manufacture a slightly modified version of Cronex 10T using Type 78 grains. TX 1341 (¶ 211); Trial Tr. at 839-43 (Hudson).

On the other hand, Cronex 10TL had always contained a blend of two different types of tabular grains. Between 1995 and early 1998, Cronex 10TL employed a blend of Type 79 (72% by weight) and Type 81 (28% by weight) grains. TX 1341 (¶ 211); Trial Tr. at 815, 839-43 (Hudson). This version of 10TL existing between 1995 and 1998 is referred to as "Cronex 10TL-old". In 1998, the grains of Cronex 10TL were replaced with a blend of Type 79LCW and Type 80 grains. The replacement version of the 10TL existing from 1998 forward is referred to as "Cronex 10TL-new". At no time did the Cronex 10T film contain any of the Type 79, 79LCW, 80, or 81 grains of Cronex 10TL. Cronex 10TL never contained the Type 88 or 78 grains of the Cronex 10T. TX 1341 (¶ 211); Trial Tr. at 839-43 (Hudson).

The lower concentration grains blended into the Cronex 10TL-old and Cronex 10TL-new emulsions were of different sizes than the grains of the Cronex 10T emulsion. The different size grains employed in the blends were added to adjust the contrast of the films. TX 832 (¶ 26); Trial Tr. 354 (Dillenbeck).

In addition to different grain types, the Type 78 grains of the Cronex 10T and the Type 79 grains of the Cronex 10TL were subjected to different emulsion precipitation procedures. TX 1136, 1341 (¶ 212); Trial Tr. 840-41 (Hudson).

The selection of different grain sizes and the different precipitation procedures between the Cronex 10T and Cronex 10TL are responsible for imparting different sensitometry properties to the respective products. TX 1341 (¶¶ 210-222), 1342 (¶¶ 8-23); Trial Tr. at 570-72 (Waaser), 681-83, 692-93 (Hurwitz), 839-43 (Hudson). Radiographic films have three primary photographic properties which are utilized by a radiologist: speed, contrast, and maximum density. TX 1342 (¶ 21). Cronex 10T and Cronex 10TL differed from one another with respect to their contrast and maximum density functionalities.

The following table illustrates some of the different properties between Cronex T, Cronex 10TL-old containing Type 79 and 81 grains, and Cronex 10TL-new containing Type 79LCW and 80 grains: TOP TOP MIDDLE UPPER FILM DENSITY GRADIENT GRADIENT GRADIENT

CRONEX 10T 3.7 1.8 3.1 2.9 CRONEX 10TL-OLD 3.3 1.4 2.4 2.0 CRONEX 10TL-NEW 3.2 1.3 2.5 2.0

TX 1060 (¶ 31). See also TX 1120, 1341 (¶ 213); Trial Tr. at 841-42 (Hudson).

Cronex 10T demonstrated a higher contrast and top density (maximum density or Dmax) than the Cronex 10TL product. Contrast describes the relative difference between light and dark areas of developed film. A high contrast film demonstrates sharp differences between the light and dark areas, whereas a low contrast film demonstrates subtle differences between the light and dark areas. The top density indicates the maximum darkness of the film. The toe gradient indicates the contrast achievable by the film at low areas of exposure. The middle gradient indicates the contrast achievable by the film at intermediate areas of exposure. The upper gradient indicates the contrast achievable by the film at areas of high exposure. TX 1120, 1341 (¶ 213), 1060 (¶ 32); Trial Tr. at 841-42 (Hudson).

Kodak's data shows substantial functional differences between Cronex 10T and Cronex 10TL. In June 1995, Kodak conducted photographic performance testing on both Cronex 10T and Cronex 10TL. TX 1246, 1341 (¶ 214). The data shows that the gradient measurements for Cronex 10TL were less at the lower, average and upper points, as compared to Cronex 10T, regardless of the screen used to expose the film. In January 2002, Kodak again studied the photographic properties of Cronex 10T and Cronex 10TL and again found photographic differences between the two films. TX 463, 1246, 1291, 1341 (¶ 214).

The functional differences between Cronex 10T and Cronex 10TL were also reflected in how Agfa and its predecessors sold and marketed the products. Cronex 10T and 10TL were sold and marketed concurrently with one another for different intended applications. TX 1341 (¶ 211); Trial Tr. at 838 (Hudson). DuPont and Sterling listed Cronex 10T and Cronex 10TL separately in their price catalogues, and DuPont and Sterling described Cronex 10T as a "high contrast" film and Cronex 10TL as a "medium latitude" film. TX 1259. DuPont, Sterling, and subsequently Agfa represented to the end user that Cronex 10T and Cronex 10TL were different products marketed for different uses based on different film properties. TX 1342 (¶ 17). Agfa markets the high contrast Cronex 10T film for applications requiring imaging of minute differences in bone structure. The lower values for the Cronex 10TL film illustrates much broader exposure latitude and is marketed for imaging the fine and dense tissues in the patient's chest region, such as lung and heart tissues. TX 1060 (¶ 32), 1120; Trial Tr. at 682-83, 685-88, 691-93 (Hurwitz), 841-42 (Hudson).

Dr. Robert Hurwitz, a diagnostic radiologist, testified that these products have been received and applied differently by radiologists. Latitude films such as Cronex 10TL provide less information about bone detail but significant diagnostic information about the adjacent soft tissues, such as the chest and abdomen, where a wide range of soft tissues and bones are encountered including the lungs, mediastinum, ribs and spine. TX 1342 (¶ 16); Trial Tr. at 682-83, 685-88, 691-93 (Hurwitz).

The testimony of Dr. Hurwitz demonstrated that radiologists appreciated the substantially different functionalities between Cronex 10T and Cronex 10TL, and that radiologists selected between these products based on how these substantial functional differences fulfilled different end user needs. Dr. Hurwitz testified a radiologist would not consider the films to be interchangeable variations of one another.

Q. And is it true that in Dr. Dillen-beck's report he gave the opinion that Cronex 10TL is merely a variation of Cronex 10T?

A. He so stated.

Q. And, Doctor, from your view point as an end user as a radiologist, do you agree with Mr. Dillenbeck's evaluation that Cronex 10TL is merely a variation of Cronex 10T?

A. I do not.

Q. And why is that?

A. Because they are two different films for two different purposes.

Q. And, Doctor, as a radiologist do you believe that these films are interchangeable?

A. I do not.

Trial Tr. at 693 (Hurwitz). See also TX 1342 (¶ 22); Trial Tr. at 682-83, 685-88, 691-93 (Hurwitz).

Kodak argues that DuPont and Sterling considered Cronex 10T and Cronex 10TL to be "sister products" and part of the same "Cronex 10T family." TX 153, 158; Trial Tr. 571 (Waaser), 1085-86 (Hudson), at 842 (Hudson); Depo. Desig. Tab 38 (10/14/05 Russell) at 168. This argument is not persuasive because had Kodak intended to make specific charges of infringement in 1995 against products other than Cronex 10T, it could have done so, just as it did in April 2001 when it charged several specific Agfa products by name with infringement. In the alternative, Kodak could have notified DuPont that the family of Cronex products violated its patents, as it did with Agfa in February 2001, when it notified Agfa that its "thoracic imaging products" violated Kodak's crossover patents.

Kodak argues that notice as to Cronex 10T constituted notice with respect to Cronex 10TL because of some shared characteristics and processing attributes of the films. Mr. Dillenbeck identified various characteristics and processing attributes for Cronex 10T and Cronex 10TL in advertising brochures that were identically or similarly worded. This argument is not persuasive because Cronex 10T and 10TL also share many identical or similar characteristics and processing attributes to Sterling products which did not use tabular grain emulsions, such as Cronex 10S. TX 125, 126, 140, 141, 1085, 1086, 1099, 1127, 1128, 1341 (¶ 217); Trial Tr. at 438-42 (Dillenbeck).

Kodak further argues that its November 10, 1995 letter to DuPont (TX 457) expanded notice of infringement beyond the Cronex 10T product. That letter made reference to technical discussions pertaining to "tabular grains in DuPont X-ray film." The November 10, 1995 letter established a confidentiality agreement between Kodak and DuPont that was to apply at the November 1995 meeting.

Q. Well, what other portion of the document mentions any other specific DuPont film that Kodak believed was infringing the `426, `425 or `520 patents that are at issue in this case?

MR. FLETCHER: Objection to the extent it goes to attorney/client privilege as to what Kodak was thinking.

A. The only film specifically mentioned is Cronex 10T.

Q. Nothing else in the letter that would provide DuPont with infringement notice other than the product of Cronex 10T, correct?

A. That's correct.

Trial Tr. at 514-15 (Hawley).

This Court has already held that:

the November 10, 1995 correspondence that Kodak contends is a notice of infringement was nothing more than a letter confirming a meeting to discuss the alleged infringement of Cronex 10T on Kodak's `425 and `426 patents. The November 10, 1995 letter itself does not charge infringement by any product of any patent, but merely confirms that the parties will meet to discuss "technical data pertaining to tabular grains in Du-Pont X-ray film and claims in Kodak's ['425 and `426] patents. . . ." The fact that the parties were meeting to discuss DuPont technology in light of Kodak's intellectual property does not demonstrate that Kodak had put DuPont on notice that all of its T-grain films infringed Kodak's patents.

Moreover, considering the November 10, 1995 correspondence in conjunction with the September 7, 1995 correspondence, as Kodak urges, only supports Agfa's contention that DuPont received notice of infringement with respect to only its Cronex 10T product. The September 1995 correspondence from Kodak to DuPont, and DuPont's October, 1995 response discuss only the Cronex 10T product. Accordingly, there is no reason to believe that Kodak's November 10, 1995 confirmatory letter was intended to expand the scope of the products at issue, given that the parties had engaged in discussions related only to the Cronex 10T film. The fact that Kodak chose to use a generic term in one confirmatory letter can not operate to transform a specific charge of infringement relating to a single product into a significantly broader charge directed at any product that may fit within that generic term.

Decision of July 2, 2004.

Rather than expanding the scope of notice from its September 1995 letter, the November 10, 1995 letter set forth an understanding that discussions at the meeting would be confidential. Even accepting, arguendo, that the November 10, 1995 letter expanded confidentiality to products other than Cronex 10T, an agreement to maintain communications confidential does not necessarily convey that all products discussed at the meeting were accused of infringement. It also bears mentioning that the confidential information exchanged at the November 1995 meeting was limited to Cronex 10T. Trial Tr. at 555 (Hawley), 584-85 (Waaser).

Considering the substantial differences between the products, the industry practice of separately designating products, Kodak's own letter of 2001 separately identifying specific products within families of products, and the other reasons discussed above, it is my recommendation that notice as to Cronex 10T did not constitute notice of infringement as to Cronex 10TL.

Defendants' Laches Defense

Defendants have asserted as a defense that Kodak delayed unnecessarily in bringing suit, so that the equitable defense of laches may be invoked. Establishing laches does not normally preclude suit, but simply limits the period for which damages may be recovered.

Laches is a judicially recognized defense to a suit for patent infringement, and Kodak's claim for damages arising prior to suit may be barred should Defendants establish entitlement to the defense. See A.C. Aukerman Co. v. R.L. Chaides Constr. Co., 960 F.2d 1020, 1028 (Fed. Cir. 1992) (en banc). Defendants allege that (1) Kodak delayed filing suit for an unreasonable and inexcusable length of time from the time that Kodak knew or reasonably should have known of its claims against Defendants; and (2) Kodak's delay operated to the prejudice or injury of Defendants. A.C. Aukerman Co., 960 F.2d at 1032; Costello v. United States, 365 U.S. 265, 282, 81 S.Ct. 534, 5 L.Ed.2d 551 (1961).

A delay of six years in filing suit from the time the plaintiff knew or should have known of its claims creates a presumption of laches. A.C. Aukerman Co., 960 F.2d at 1035-36. The parties dispute whether a period of six years passed since Kodak knew or should have known of its claims of infringement.

The following is generally undisputed. Kodak sent a September 1995 letter to DuPont (predecessor to Defendants) asserting that DuPont's Cronex 10T product infringed the `425 and `426 patents. TX 150. After an exchange of correspondence between Kodak and DuPont, a meeting took place in North Carolina on November 18, 1995. TX 828 (¶ 3); Trial Tr. 502 (Hawley). DuPont claimed that it was experiencing start-up difficulties at its then new Brevard facility which may have resulted in a production run tested by Kodak that was out of specification. TX 1046; Trial Tr. at 504, 526-27, 557 (Hawley), 574-78, 581-82 (Waaser), 858-62 (Hudson); Dep. Tr. of Apple 1/29/04 at 25-30, Dep. Tr. of Guy 2/6/04 at 107-109, Dep. Tr. of Braitsch 9/9/04 (30(b)(6)) at 38-40.

The meeting concluded by DuPont assuring Kodak that DuPont intended to operate comfortably outside the claims of the T-grain patents, and that any excursions due to "start-up" problems were "fixed." TX 828 (¶ 4); Trial Tr. 508-09 (Hawley); Depo. Desig. Tab 2 (Apple) at 220. Consistent with the representations at the meeting, DuPont agreed to provide Kodak with film codes for Cronex 10T film produced after the start-up issues at its Brevard facility had been corrected so that Kodak could test to assure that DuPont's assertions of non-infringement were correct. TX 1046; Trial Tr. at 160 (Van Dam), 529-530, 552-553 (Hawley), 581 (Waaser), 858-62 (Hudson).

DuPont provided the promised product codes to Kodak in November 1995. DuPont advised Kodak that: "Further, we understand that Kodak will purchase publicly available Cronex 10T with an emulsion number of 578-511x-0161, or later, to confirm that the product is not covered by either of the above identified patents. We assume that once Kodak is satisfied that our product is not covered by either patent, Kodak will drop its charge of infringement." TX 433, 1256; Trial Tr. at 557 (Hawley), 584-85 (Waaser).

In May 1996, Kodak performed grain size measurements on a sample of Cronex 10T that had been made after the batch identified in DuPont's letter of November 29, 1995. Kodak thereafter analyzed the data from those measurements to compare Cronex 10T to the patent claims. Kodak did not notify DuPont or Sterling of any allegations of infringement as a result of this testing. TX 391, 1258, 828 (¶ 7), 1049 (¶ 7); Trial Tr. at 561 (Hawley), 587 (Waaser); Dep. Tr. of Braitsch 9/9/04 (30(b)(6)) at 39-80.

Q. And did Kodak, in fact, analyze a subsequent batch after the 1995 discussions between DuPont and Kodak regarding the alleged infringement by Cronex 10T?

A. Kodak did analyze a subsequent batch.

Dep. Tr. of Braitsch 9/9/04 (30(b)(6)) at 39.

Q. Did the business at Eastman Kodak responsible for making the decision in 1996 whether to sue DuPont over the alleged infringement of Cronex 10T have the subsequent analysis in 1996 of Cronex 10T?

MR. NOTO: Objection to form.

A. I can't personally say that they had that information. I was told that they had that information.

Dep. Tr. of Braitsch 9/9/04 (30(b)(6)) at 41.

The grain thicknesses, average aspect ratio, individual projected area and percent total projected area for the Cronex 10T product (578 6011 0070 01) analyzed in 1996 were calculated by Agfa based on Kodak's data contained in TX 1258. The percent total projected area was calculated from the sum of the projected areas of tabular grains having a thickness of less than 0.2 micron divided by the sum of the projected areas of all grains measured. From Kodak's data, the average aspect ratio of this subset of grains was 7.04 for those grains having a thickness of less than 0.2 micron and the percent total projected area was 21.10 percent. TX 1060 (i 22). Defendants have represented that these 1996 measurements resulted in at least 50 percent of the total projected area being provided by grains having a thickness of 0.233 micron or less, with the grains having an average aspect ratio of 6.30:1. TX 1060 (¶ 24).

Kodak asserts infringement under the doctrine of equivalents for various products where grains of up to 0.246 micron are included to reach the claimed limitation of at least 50% of total projected area. Other products having an average aspect ratio as low as 4.4:1 are also alleged to infringe Kodak's T-grain patents under the doctrine of equivalents. The calculation for the 1996 Cronex 10T film sample falls within the thickness and aspect ratioranges claimed by Kodak to infringe claim 1 of the `426 patent under equivalency. TX 832 (¶ 14). Indeed, Kodak has not denied that its claim scope under the doctrine of equivalents encompasses the 1996 Cronex 10T film. See TX 1053 (No. 219).

A period of more than six years elapsed between Kodak's May 1996 testing of the Cronex 10T product and the filing of this action on October 31, 2002. The breadth of the claim scope for which Kodak alleges infringement under the doctrine of equivalents is sufficiently broad to cover the Cronex 10T tested more than six years before suit was filed. Defendants have therefore demonstrated that, with respect to Cronex 10T, there was a delay of at least six years in filing suit from the time that Kodak knew or should have known of its claim against Cronex 10T. Accordingly, a presumption of laches applies against Kodak with respect to Cronex 10T.

An effect of this presumption is to shift the burden onto Kodak to come forward with sufficient evidence to raise a genuine factual issue regarding the reasonableness of the delay or the absence of prejudice. A.C. Aukerman Co., 960 F.2d at 1037-39. The presumption is of the "double bursting bubble" type in the sense that if Kodak raises a genuine issue of fact with regard to either element, then Agfa must carry the burden of persuasion and affirmatively prove both elements in order to prevail on the defense of laches. Hemstreet v. Computer Entry Sys. Corp., 972 F.2d 1290, 1293 (Fed. Cir. 1992). As explained below, Kodak has demonstrated sufficient excuse for its delay to rebut the presumption. Agfa, on the other hand, has not carried its burden of persuasion sufficient to establish entitlement to the defense.

Following the November 1995 meeting and its subsequent testing, Kodak made no further allegations of infringement of the T-grain patents to DuPont or its immediate successor Sterling. Trial Tr. at 561 (Hawley), 587, 596 (Waaser), 639 (Verhoeven); Dep. Tr. of Braitsch 9/9/04 (30(b)(6)) at 166-83. The evidence indicates that Kodak's testing of the 1996 Cronex 10T film in May of 1996 may have provided Kodak with sufficient information to form at least a preliminary opinion that the tested batch of film infringed Kodak's patents under Kodak's interpretation of its claims under the doctrine of equivalents. Kodak had accepted DuPont's explanation and assurances made during the November 1995 meeting. TX 828 (03/02/04 Hawley declaration, ¶ 8); Trial Tr. 509 (Hawley). However, Kodak relied on DuPont's representations and assurances that it would permanently fix the manufacturing problems. TX 828 (03/02/04 Hawley declaration ¶ 8); Trial Tr. 509 (Hawley). Kodak's reliance was not unjustified. Kodak believed DuPont to be a very reputable company. Indeed, the initial samples tested supported DuPont's representations that the start-up problems had been resolved.

Kodak monitored the Cronex 10T film through at least 1998. TX 469, 1261; Trial Tr. at 160 (Van Dam); Dep. Tr. of Braitsch 9/9/04 (30(b)(6)) at 132-33, 149. In 1998, Kodak again analyzed a Cronex 10T product sample for grain sizes. TX 1261. As explained by David Braitsch, Kodak's Intellectual Property Manager for research and development at the time, routine benchmark testing at Kodak in mid-1998 revealed a decrease in silver content in film manufactured by Sterling. TX 829 (03/03/04 Braitsch declaration ¶ 5); Agfa Depo. Desig. Tab 4 (09/09/04 Braitsch) at 31-32. The testing served to arouse Kodak's concerns.

Once Kodak's concerns were aroused, it investigated the Cronex 10T product and then expanded that investigation. It would seem that Kodak acted reasonably in investigating other products once its concerns about Cronex 10T were aroused.³ While Kodak could have notified Sterling at the time about the product it then examined, it may well have believed another "manufacturing" issue had arisen. Agfa Depo. Desig. Tab 4 (09/09/04 Braitsch) at 117. Further, Kodak's individual grain-by-grain thickness analysis of the medical imaging film manufactured by Sterling was necessary for determining infringement under the Kodak T-grain patents. TX 829 (03/03/04 Braitsch declaration ¶ 5). Testing of the other products in suit extended the amount of time required for Kodak to complete its investigation. It was reasonable for Kodak to carry out these tests and, as both parties have attested to, the shadowed electron microscopy technique is time consuming. Trial Tr. 1178 (Liggero) (process is "labor intensive"); Agfa Depo. Desig. Tab 4 (09/09/04 Braitsch) at 136-37. It was also reasonable for Kodak to seek advice of legal counsel before making accusations of infringement against Agfa. Kodak took the time reasonable under the circumstances to review, analyze, and evaluate the Sterling and Agfa T-grain film products in depth before notifying Agfa of its claims for infringement.

3 This should not be construed as implying that Kodak's equivalency claims are meritorious.

Agfa's actions also demonstrate the time commitment involved in conducting an infringement analysis of the T-grain patents. After Kodak sent Agfa a letter regarding infringement of the T-grain patents in April 2001 (less than six years after its original September 1995 notice letter to DuPont regarding Cronex 10T), Agfa by its own admission required a relatively long time to conduct internal discussions regarding the patent infringement allegations due to the complexities of the issues, the need for product testing, collecting of relevant information, and subsequent technical and legal analysis. Trial Tr. 750 (Goedeweeck). It initially took Agfa almost three months to respond to Kodak's letter regarding the T-grain patents. TX 352; Trial Tr. 749 (Goedeweeck). Agfa continued to request further delays before substantively responding, arguing that it was "premature" to have further discussions in July 2001. Trial Tr. 758 (Goedeweeck). See also TX 241.

Kodak first informed Defendants that it believed that Sterling Ultravision Ci, Sterling Ultravision L, Sterling Ultravision C, Agfa Cronex 5, Agfa B Plus, Agfa CP-BU, Agfa Ortho HT-G, Sterling Cronex 10T, Sterling Orthovision L, Sterling Orthovision G, and Agfa Ortho Opthos H infringed at least one or more of the T-grain patents in an April 9, 2001 letter. TX 432, 1281. In response to Kodak's April 2001 infringement allegations, Defendants informed Kodak that, according to their assessment, none of the products or emulsions were covered by any of the three patents. TX 352, 1282; Trial Tr. at 720-23 (Goedeweeck). As described above, it took Defendants several months to complete their assessment. On September 6, 2001, Agfa and Kodak exchanged technical data on Opthos H, one of the accused products. TX 458, 1165, 1284; Trial Tr. at 726-29 (Goedeweeck).

On October 29, 2001, Kodak terminated discussions with Defendants. Despite a tradition of handling disputes outside of court, Kodak did not contact Defendants again until over a year later when Kodak filed suit. TX 376, 1288; Trial Tr. at 734-36 (Goedeweeck).

While Kodak conceivably could have proceeded more expeditiously after its concerns were aroused, Defendants suggested a product-by-product evaluation during which the parties could air their positions. Hence, Agfa itself was proposing a protracted process to resolve matters. When the 2001 discussions ended, Kodak proceeded to engage counsel and prepare for litigation. Hence, it appears that Defendants seek the benefit of the failed discussions and their own delay, while faulting Kodak for not initially suing in 1996, shortly thereafter, or prior to sending its 2001 letter.

Furthermore, this suit involves products other than those produced by Du-Pont/Sterling. Various Agfa products have been accused of infringement. In short, Agfa would seem to be seeking a piecemeal litigation process which would be overly complicated by requiring that each product's initial awareness by Kodak be investigated. There is no suggestion that Kodak had significant information about products other than Cronex 10T in 1996, at least not information sufficient to formulate an accusation of infringement against those other products. If Agfa were to have its way, Kodak would be required to file an infringement action after it learned of infringement by any one Agfa product, and to thereafter supplement its complaint every time Kodak initially became aware that an accused product infringes a T-grain patent. This would lead to piecemeal litigation, repetitious discovery and motion practices, and other complications.

Moreover, Agfa has not established prejudice or injury. Prejudice can be evidentiary or economic. A.C. Aukerman, 960 F.2d at 1034. While Agfa established that certain participants at the 1995 meeting had in the interim died, it did not establish any meaningful evidentiary prejudice as a result because other participants either testified or gave depositions. Although Dr. Verhoeven, Chairman of Agfa's Board of Directors, testified that the Sterling acquisition either would not have gone forward or would have been structured differently had he been aware of the infringement issue, he was aware of the issue because the 1995 letter was produced during due diligence in the course of the Sterling acquisition. Despite that, he did not investigate the matter further. Hence, Agfa has not demonstrated any meaningful economic prejudice in its acquisition of Sterling. Additionally, the products in suit involve also Agfa products and not just products produced by Sterling.

Finally, laches is an equitable remedy. A laches defense may be defeated if the plaintiff can show that the defendant "has engaged in particularly egregious conduct which would change the equities significantly in plaintiff's favor." Id. at 1033. While the claims contain limitations other than mere grain thickness, DuPont and Sterling never informed Kodak at or after the 1995 meeting of the to shift to a thinner grain. Potash Co. of Am. v. Int'l Minerals Chem. Corp., 213 F.2d 153, 155 (10th Cir. 1954) (if "the party which advances the defense of laches is responsible for the delay or contributes substantially to it he cannot take advantage of it. Then, too, if material facts are concealed or misrepresented by a suspected wrongdoer, and if in reliance thereon the person wronged is deceived and his suspicions are allayed for awhile, a court of equity will not grant the wrongdoer any advantage resulting from the elapse of time").

The absence of an opinion of counsel is a factor to be considered in determining the appropriateness of laches. Bound v. Spencer Gifts, Inc., 1996 WL 556657, 1996 U.S. Dist. LEXIS 14325 (E.D.Pa. 1996) ("a defendant's failure to make a reasonable inquiry regarding patent validity after being apprised of its existence is a factor to be considered" in determining whether the conduct was egregious to defeat a laches defense); Cover v. Hydramatic Packing Co., 34 U.S.P.Q.2d 1128 (E.D.Pa. 1994) (defendant found to have engaged in egregious conduct by continuing to manufacture a product after receiving an infringement notice and without conducting a reasonable inquiry into the validity of the patents). Dr. Verhoeven testified that the 1995 notice letter from Kodak to DuPont was produced during the due diligence phase of Agfa's acquisition of Sterling, which had by then been spun-off from DuPont. Trial. Tr. at 641. Although he testified that he had assumed the matter concluded by the time of the acquisition, he did not testify to having received any advice from U.S. counsel on the matter. Trial. Tr. at 644. To the extent that the equities were not already balanced in Kodak's favor by the time required to investigate the various products and the lack of prejudice, Defendants' failure to obtain legal counsel and their silence concerning the product changes shift the equities in Kodak's favor.

In view of the above, it is recommended that Defendants have not established entitlement to the defense of laches. Defendants' Invalidity Challenges

As in essentially every patent case, the accused infringers assert that the allegedly infringed claims are invalid in view of prior art.⁴ As the party asserting invalidity, Defendants* burden is one of clear and convincing evidence. 35 U.S.C. § 282. "A patent issued from the United States Patent and Trademark Office (PTO) bears the presumption of validity under 35 U.S.C. § 282. An accused infringer, therefore, must prove patent invalidity under the clear and convincing evidentiary standard." Metabolite Labs., Inc. v. Laoratory. Corp. of Am. Holdings, 370 F.3d 1354, 1365 (Fed. Cir. 2004), cert. denied, 548 U.S. 926, 126 S.Ct. 2976, 165 L.Ed.2d 990 (2006). Moreover, it is "especially difficult" to satisfy the burden of proving invalidity when prior art cited in litigation was previously considered by the PTO during prosecution. Hewlett-Packard Co. v. Bausch Lomb, Inc., 909 F.2d 1464, 1467 (Fed. Cir. 1990).

4 The inequitable conduct defense will be addressed separately.

Prior art may render claims invalid as "anticipated" under 35 U.S.C. § 102 or as "obvious" under 35 U.S.C. § 103. Anticipation requires that each and every element of the claimed invention be disclosed in a single prior art reference. In re Paulsen, 30 F.3d 1475, 1478-79 (Fed. Cir. 1994). Obviousness requires that "the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains." 35 U.S.C. § 103. "[A] patent composed of several elements is not proved obvious merely by demonstrating that each of its elements was, independently, known in the prior art." KSR Int'l Co. v. Teleflex Inc., ___ U.S. ___, 127 S.Ct. 1727, 1741, 167 L.Ed.2d 705 (2007). "[I]nventions in most, if not all, instances rely upon building blocks long since uncovered, and claimed discoveries almost of necessity will be combinations of what, in some sense, is already known." KSR, 127 S.Ct. at 1741.

Kodak sold a commercial X-ray film product containing chemically sensitized, tabular grain emulsions since the 1930's.

From 1937 until the 1950's the Eastman Kodak Company sold a Duplitized(r) radiographic film product under the name No-Screen X-Ray Code 5133. The product contained as coatings on opposite major faces of a film support sulfur sensitized silver bromide emulsions. Since the emulsions were intended to be exposed by X-radiation, they were not spectrally sensitized. The tabular grains had an average aspect ratio in the intermediate range of from about 5 to 7:1. The tabular grains accounted for greater than 50% of the projected area while nontabular grains accounted for greater than 25% of the projected area.

TX 312 (`520 patent, col. 8, l. 66, to col. 9, l. 15); TX 320 (`425 patent, col. 1, l. 62, to col. 2, l. 12); TX 313 (`426 patent, col. 1, l. 62, to col. 2, l. 13); Trial Tr. at 1101 (Liggero).

One cannot conclude from the foregoing statement that the No-Screen X-Ray Code 5133 emulsions met the grain structure and other parameters of either the intermediate or the high aspect ratio claims. TX 835 (10/26/06 Marchetti report, ¶ 36); Trial Tr. 1147 (Liggero) (none of the claims of the patents in suit call for an emulsion with an average grain thickness of 0.36 micron which is a characteristic of No-Screen according to the T-grain patents).

Further, No-Screen was not spectrally sensitized. TX 835 (10/26/06 Marchetti report, ¶ 37). See also Trial Tr. 1146 (Liggero). Nothing about No-Screen would have prompted or encouraged a person skilled in the art to investigate spectral sensitization of tabular grains. TX 835 (10/26/06 Marchetti report, ¶ 37). As the above text indicates, No-Screen was not spectrally sensitized because it was intended to be exposed to X-ray-radiation (not to light from intensifying screens, as is the case for the films at issue in this litigation). Id; Trial Tr. 1146 (Liggero).

Moreover, Kodak specifically referred to the No-Screen product in the patent applications and brought the product to the attention of the patent examiner in an Information Disclosure Statement. TX 1009 (at EK027144). It is "especially difficult" to satisfy the burden of proving invalidity when prior art cited in litigation was previously considered by the PTO during prosecution. Hewlett-Packard, 909 F.2d at 1467.

C. Mees T. James, The Theory of the Photographic Process (3d ed. 1966), Figure 2.6 ("Mees and James") is widely accepted as the authoritative work of the photographic sciences. "The Mees and James book . . . is widely read and in many ways the Bible of photographic science." TX 1025; Trial Tr. at 1108, 1111 (Liggero).

Mees James Fig. 2.6 is described as showing "[p]reshadowed carbon replicas of grains of a high-speed negative-type emulsion." TX 846 (Mees James at 37). It is a single-field image. TX 835 (10/26/06 Marchetti report, ¶ 38). The text referring to this Figure states that "[t]he grains of commercial emulsions (Figs. 2.5 and 2.6) are rarely larger than 5; a more normal upper limit for very fast emulsions is half that value." TX 846 (Mees James at 36).

Using the disclosed shadow angle of 11 degrees and ignoring sampling and measurement error, Agfa's expert Dr. Liggero measured 42 grains in Figure 2.6 and concluded that the depicted emulsion contains tabular grains of less than 0.3 micron in thickness and at least 0.6 micron in diameter that account for about 89 percent of the total projected area and have an average aspect ratio of about 11.5:1. TX 1025A, 1169; Trial Tr. at 1110-11 (Liggero).

Dr. Liggero's measurement protocol and the photograph of Figure 2.6 of Mees James are not sufficiently reliable as to constitute clear and convincing evidence of invalidity.

First, the number of grains measured by Dr. Liggero is insufficient to clearly and convincingly constitute a reliable sampling population for determining whether the Mees and James emulsion has a high aspect ratio as required by the `425 and `520 patents or an intermediate aspect ratio as defined by the `426 patent. According to a document generated by an Agfa affiliate in 1987, only 30 of the grains in this image can be measured. TX 76 ("The micrograph . . . has several crystals that cannot be measured because they overlap leaving 30 crystals in the field"); Trial Tr. at 1149-50 (Liggero). Despite that, Dr. Liggero testified that he was able to measure as many as 42 grains. Trial Tr. at 1110 (Liggero); Depo. Desig. Tab 31 (02/13/07 Liggero) at 164. Dr. Liggero himself stated, however, that 42 grains is an insufficient number of grains from which to draw a conclusion about whether the emulsion falls within the literal scope of any of the T-grain patents. Trial Tr. 1110, 1150, 1200 (Liggero) ("To draw rigorous scientifically accurate conclusions, one would need more than 30 or 42 samples. . . .").

Second, the quality of the photograph in Mees and James draws Dr. Liggero's measurements into question. According to Dr. Liggero, to "draw rigorous scientifically accurate conclusions, . . . one would want the original and standard TEM." Trial Tr. 1200. Dr. Liggero also testified that he was not sure whether the image some-how had been changed from the original as a result of its being reproduced for bookmaking purposes. Trial. Tr. 1198. Dr. Liggero indicated that his measurements from the textbook were sure to include distortion error. Trial Tr. 1111.

Third, Dr. Liggero's conclusions that the emulsion shown in Figure 2.6 of the Mees and James reference was chemically and spectrally sensitized were based on text that the emulsion was a commercial emulsion. Dr. Liggero stated that the emulsion was "most likely optimally chemically and spectrally sensitized." Trial Tr. at 1108-11 (Liggero). This testimony concerning "likely" optimum sensitization does not rise to the level of confidence necessary for satisfying a clear and convincing burden. Recall that No-Screen, a tabular grain product, was commercially available and yet was not spectrally sensitized.

Kodak's expert witness Dr. Marchetti asserted that Mees James Fig. 2.6 does not teach anything about sensitization of T-grains. TX 835 (10/26/06 Marchetti report, ¶ 40). Kodak represented during the reexamination proceedings that the emulsion shown in Mees James Fig. 2.6 was a research emulsion and was not spectrally sensitized. TX 423 (affidavit submitted by Kodak in the `426 reexamination). Dr. Liggero stated that he had no reason to doubt the accuracy of Kodak's statements. Trial Tr. 1187-88 ("I have no reason to believe that Kodak's statement to the Patent Office in the `426 re-exam wasn't accurate.").

Finally, the `426 patent was subjected to reexamination by the PTO. Mees James Fig. 2.6 was before the PTO in the Reexamination of the `426 patent. Trial Tr. 1147-48; Depo. Desig. Tab 31 (02/13/07 Liggero) at 142. The statutory "presumption [of validity] is fortified to the extent that there has been a reexamination of the patent by the Patent and Trademark Office." CVI/Beta Ventures, Inc. v. Custom Optical Frames, Inc., 893 F.Supp. 508, 514 (D.Md. 1995), aff'd, 92 F.3d 1203, 1996 WL 338388 (Fed. Cir. 1996); E.I. du Pont de Nemours Co. v. Polaroid Graphics Imaging, Inc., 706 F.Supp. 1135, 1141 (D.Del. 1989) (while presumption of validity remains unaltered, "the exhaustive consideration given the prior art by the PTO during [reexamination] must be weighed in determining patentability"), aff'd, 887 F.2d 1095, 1989 WL 108217 (Fed. Cir. 1989).

The James and Higgins reference in Figure 2.3 depicts a "typical negative-type" silver iodobromide tabular grain emulsion. According to Agfa, the reference therefore describes to a person of ordinary skill in the art a commercial optimally chemically and spectrally sensitized tabular grain film. TX 1021A; Trial Tr. at 1114-16 (Liggero). Agfa further argues that using the disclosed shadow angle of 18.4 degrees and ignoring sampling and measurement error, Figure 2.3 depicts an emulsion containing intermediate aspect ratio T-grains as defined by the `426 patent claims.

During cross-examination Dr. Liggero acknowledged that Fig. 2.3 of James and Higgins was prepared from the same electron micrograph as Figure 2.6 of Mees and James. Trial. Tr. 1165 (Liggero). Dr. Liggero also stated that he could not find any reason to doubt Kodak's statement to the PTO that the sample was mislabeled and was not a commercial sample. Trial Tr. 1188 (Liggero). Clear and convincing evidence requires more than assumption. As the party asserting invalidity, it was incumbent upon Defendants to provide evidence that was both clear and convincing about the James and Higgins material. Defendants have not met this burden.

G. Danguy, "On the Sensitometric Character and Chemical Sensitization of Silver Bromide Emulsions with Tabular Grains," Science et Industries Photographiques, Vol. 34, No. 5-6, at 141-50 (1963) (TX 1024) ("Danguy 1963"), was cited by Kodak for consideration by the PTO during prosecution of the `520 patent. See TX 321 (EK004997-5005), 1009 (EK027143-46) (Cited Art Statement Under 37 CFR 1.97-1.99). Kodak characterized Danguy as follows:

Danguy can be viewed as being essentially cumulative with de Cugnac and Chateau, cited in the specification at page 2, lines 21-26, in that it also discloses high aspect ratio tabular grain silver bromoiodide emulsions. Danguy further discloses chemical sensitization of high aspect ratio tabular grain emulsions, which de Cugnac and Chateau does not.

TX 321 (EK005000), 1009 (EK027146).

Danguy 1963 is an elaboration of a study first reported in G. Danguy, "The Preparation and the Sensitometry of Photographic Emulsions with Tabular Silver Bromide Grains," Bulletin de la Societé Rovale des Sciences de Liège, Vol. 31, No. 11-12, at 732-50 (1962) (TX 364) ("Danguy 1962"). See TX 835 (10/26/06 Marchetti report, ¶ 12).

Danguy 1962 and Danguy 1963 describe the preparation of five emulsions, each precipitated according to the same conditions and then ripened for progressively longer periods. TX 835 (10/26/06 Marchetti report, ¶ 13). Of the five emulsions, the fifth emulsion (the "845" emulsion) was subjected to 15 hours of ripening, by which time "all the small initial grains [had] disappeared." TX 1024 (at 143). The emulsion was described as being "formed exclusively of tabular grains of greatly varying dimensions," including a thickness "on the order of 0.1µ." Id. The "large crystals" of the emulsion are described as reaching 50 µ, with a few crystals even larger. Id. The grains are described as having linear dimensions of between 25 and 100. According to Danguy 1962, the average size of the grains is "about 30 and can be as high as 100 µ". TX 364.

Kodak stated in its disclosure of Danguy 1963 to the PTO that these statements indicate the presence of high aspect ratio grains, and, in that sense, they are cumulative of de Cugnac. TX 835 (10/26/06 Marchetti report, ¶ 14); TX 321 (at EK005000), 1009 (at EK027146). Although, as admitted by Kodak, Danguy discloses an emulsion having high aspect ratio grains, Kodak's T-grain patents do not claim their inventions so broadly. Ekchian v. Home Depot, Inc., 104 F.3d 1299, 1304 (Fed. Cir. 1997) ("statements made in an IDS can be the basis for a court to interpret the scope of the claims of a granted patent.") Rather, the T-grain patent claims include specific limitations concerning the thickness, average aspect ratios and percent projected areas of the grain populations. Danguy does not provide sufficient information to determine whether the described emulsion grain populations meet the grain structure and parameters of either the high aspect ratio claims or the intermediate aspect ratio claims. TX 835 (¶ 14). The grain precipitation and ripening processes described in Danguy 1963 (at p. 143) would not inherently or necessarily result in emulsions meeting these parameters. Id.

Additionally, Defendants have not demonstrated by clear and convincing evidence that Danguy discloses spectral sensitization or optimal spectral sensitization of its tabular grains. Danguy 1963 reports having chemically sensitized emulsion "845" (the fifth emulsion described above) to provide emulsion "937". TX 321, 835 (10/26/06 Marchetti report, ¶ 16), 1024 (at 148). Agfa concludes that the chemical sensitization disclosed in Danguy 1963 is optimal chemical sensitization of the high aspect ratio tabular grains. Agfa bases its conclusion on Danguy's statement that "[t]he decrease in concentration of Au + S complex is necessitated by the limitation of fogging of these two emulsions, optimally sensitized as in the previous case." TX 1024, at 148; Trial Tr. at 1107 (Liggero).

However, neither Danguy 1962 nor Danguy 1963 disclosed spectral sensitization (optimal or otherwise) of the disclosed tabular emulsions. TX 835 (10/26/06 Marchetti report, ¶ 20). In light of the results reported in Danguy 1963, a person skilled in the art would have been discouraged from attempting to spectrally sensitize an emulsion containing tabular grains like those disclosed in that article. TX 835 (10/26/06 Marchetti Report, ¶ 19). Dr. Marchetti's expert report, largely unrebutted, advises that the relatively high volume tabular grains of Danguy 1963 exhibited only comparable native sensitivity to much smaller volume non-tabular (globular) grains discussed in Danguy. See TX 835 (10/26/06 Marchetti report, ¶¶ 15-20). One would have expected the higher volume tabular grains to have exhibited much greater native sensitivity. Id.

Finally, Danguy 1963 states that the emulsions were exposed to light filtered by a Wratten Kodak blue filter light no. 50. TX 1024 (at p. 148), 835 (10/26/06 Marchetti report, ¶ 17). Kodak argues that by this technique, the emulsions were exposed solely to blue light, in order to evaluate speed in the native region of sensitivity. Id.

Danguy 1963 refers to a commercial emulsion as having been spectrally sensitized. TX 835 (10/26/06 Marchetti report, ¶ 21), 1024 (at 148). Kodak argues that this reference in Danguy 1963 is not a teaching of spectral sensitization of tabular grains. TX 835 (¶ 21). The commercial emulsion discussed in Danguy 1963 contained non-tabular grains. Id. Further, the spectral sensitization of the commercial emulsion played no role in the analysis presented in Danguy 1963. Id. The comparative experiment described in Danguy 1963 — involving exposure to light filtered by a Wratten Kodak blue filter light no. 50 — was designed to eliminate any effect on the sensitization comparison resulting from the presence of a green spectral sensitizer in the commercial emulsion. Id.

Hence, while Danguy may have disclosed high aspect tabular grains, the patents do not claim the inventions so broadly. Indeed, while Kodak acknowledged that tabular grains were known before the invention but characterized them as scientific curiosities of no commercial or practical application, the T-grain patents acknowledge that the No-Screen product was a tabular grain X-ray imaging product⁵. Kodak's characterization of tabular grains as primarily a scientific curiosity is corroborated by Dr. Liggero, Defendants' expert, who testified that while at Polaroid he was aware of tabular grains but recalled no investigations or research that suggested that the tabular grains would be useful in Polaroid's film products. Trial Tr. at 1100 (Liggero) ("In my early years at Polaroid tabular grain emulsions were considered for various projects, but rejected"), 1145 (no one at Polaroid in the 1970s came up with tabular grains that would provide a benefit)

5 The No-Screen product was not spectrally sensitized, however.

Between 1971 and 1974, Fuji Film Co., Ltd. ("Fuji") scientist Dr. Tanaka Tani authored a nine part series concerning studies on the spectral sensitizing effect of dyes on tabular grain emulsions. TXs 1027-1034; Trial Tr. at 1117-19 (Liggero). The ninth paper in that series, T. Tani, "Photographic Effects of Electron and Positive Hole Traps in Silver Halides DC. Factors Influencing Desensitization Caused by Sensitizing and Desensitizing Dyes," Photographic Science Eng'g, Vol. 18, at 576-81 (Sept-Oct. 1974) (TX 1034) ("Tani IX"), investigated desensitization of emulsions to blue light resulting from application of a known visible-range sensitizing dye (referred to as "Dye I" in the article) and a known desensitizing dye (referred to as "Dye II" in the article). TX 835 (10/26/06 Marchetti report, ¶ 22).

Agfa argues that Tani EX describes tabular grains having an average thickness of 0.2 micron and an average grain diameter of 1.11 micron — i.e., an average aspect ratio of 5.5:1. TX 1034, 576-77; Trial Tr. at 1118-19 (Liggero). Agfa also points to Figure 1 of Tani DC as depicting an emulsion consisting of predominately thin tabular grains — i.e., at least 50 percent of the total grain projected area consists of grains of less than 0.2 micron in thickness. TX 1034; Trial Tr. at 1118-19 (Liggero). However, in Tani IX the phrase "height of 0.2 |x" is preceded by the words "under the assumption that." In this passage, Tani IX seemingly is assuming a grain height to estimate the amount of Dye I required for monolayer coverage of the grains. Id.; see also Trial Tr. at 1156-57 (Liggero); Depo. Desig. Tab 31 (02/13/07 Liggero) at 23 ("Q. So Professor Tani is assuming those grain dimensions in order to deduce the amount of dye necessary for a mono-layer coverage of the grains, correct? MR. LOWE: Objection to form. BY THE WITNESS: Yes").

Tani IX does not provide sufficient information to determine whether the Type A emulsion met the grain structure and parameters of either the intermediate or the high aspect ratio claims. TX 835 (10/26/06 Marchetti report, ¶ 26). Nothing in the reference discloses that actual measurements were conducted to determine whether greater than 50% of the total projected area was comprised of grains with a thickness less than 0.2 micron. Trial Tr. 1157 (Liggero). Nothing in the reference discloses that actual measurements were conducted to determine whether greater than 50% of the total projected area was comprised of grains with a thickness less than 0.3 micron and a diameter of at least 0.6. µ Trial Tr. 1157 (Liggero). The portions of Tani IX relied upon by Agfa state assumptions.

Drs. Liggero and Hudson attempted to follow the grain production recipe described in Tani IX, producing an emulsion wherein tabular grains of less than 0.3 micron in thickness and at least 0.6 micron in diameter provided about 64 percent of the total grain projected area and had an average aspect ratio of about 7.2:1. TX 1334 at ¶¶ 21-22. Kodak responds that Agfa's purported "replication" of a Type A emulsion involves a number of ingredients, volumes, and steps not disclosed in Tani DC. TX 835 (10/26/06 Marchetti report, 26). For example, the gel concentration is not specified in Tani DC Id. Gel concentration is a significant parameter in determining size and thickness. Id.; see, e.g., TX 856 (M. Antoniades J.S. Wey, "Aggregation Phenomena in AgX Precipitation in the Presence of Gelatin," J. Imaging Science Tech. 42:393-98 (1998)), 853 (M. Antoniades J.S. Wey, "Precipitation of Fine AgBr Crystals in a Continuous Reactor: Effect of Gelatin on Agglomeration," J. Imaging Science Tech. 36:517-24 (1992)). Thus, it is impossible to "replicate" the Type A emulsion based on the disclosures in Tani DC, and Agfa's attempted "replication" provides insufficient basis for arriving at clear and convincing conclusions about the grain parameter characteristics of the Type A emulsion. It may be noted that if this replication could be deemed accurate and reliable, it would conflict with Agfa's position that Tani IX's emulsion has a thickness less than 0.2µ under the `426 patent claims.

With regard to spectral sensitization, Agfa argues that Tani DC teaches the spectral sensitization of the tabular grains using a green spectral ("visible range") sensitizing dye named 3,3'-diethyl-9-methyl-thiacarbocyanine bromide. TX 1034, 576; Trial Tr. at 1118-20 (Liggero). According to the results reported about Figure 2, Tani IX states that "the examination on the desensitization by the dyes has been extended over ranges from the low to the high level of the dyes, which exceeded their optimum amounts." TX 1034 at 577. However, the accompanying Figure 2 of Tani IX does not show sensitization at any level of Dye I or Dye II. Rather, Figure 2 shows desensitization as a function of the amount of dye added. As acknowledged by Dr. Liggero, Defendants' expert, desensitization is not the same as sensitization. Trial Tr. at 1145. Hence, Tani IX investigated a mechanism different than sensitization. While the two are seemingly related, the effect of one is not necessarily the effect achieved in the other. Therefore, "optimum amount," as used by Tani IX, should be interpreted simply to mean assumed monolayer coverage. TX 835 (10/26/06 Marchetti report, ¶ 29).

Finally, the non-pertinence of Tani IX is suggested by the actions of Fuji, another competitor. Fuji and Agfa communicated about the Kodak T-grain patents and the possibility of challenging them with prior art. TX 932. Significantly, although Fuji provided Agfa with a listing of prior art that it deemed potentially pertinent to the patents, Fuji did not identify the Tani articles. Considering that Dr. Tani was a Fuji scientist, it would have been expected that Fuji would have cited those articles if they had been deemed important. Fuji was apparently aware that publications could be cited as prior art, because its letter references two publications. TX 932. Furthermore, the Tani article predates the filing of the T-grain patents by almost ten years and there is nothing in the record that Fuji commercialized, attempted to commercialize or recognized the Tani articles as teaching what Defendants claim is taught. Considering that Fuji and Kodak were fierce competitors in the film business, it is difficult to believe that Tani disclosed such a significant breakthrough that nonetheless was ignored by Fuji.

Agfa argues that Figure 11 of Tani III (reference 1(c) in Tani IX) [T. Tani, "Photographic Effects of Electron and Positive Hole Traps in Silver Halides III. Desensitization by Sensitizing and Desensitizing Dyes," Photographic Science Eng'g, Vol. 15, No. 5, at 384-393 (Sept.-Oct. 1971) ("Tani III")] shows the addition of the green spectral sensitizing dye to the tabular grains described in Tani IX in amounts sufficient to achieve a maximum speed attainable from the grains in the "minus blue" ( i.e., green) region of the spectrum. TX 1029 (Figure 11 and accompanying text), 1336; Trial Tr. at 1120-21 (Liggero).

Tani III does not disclose chemical sensitization of T-grains. Dr. Liggero conceded as much on cross-examination. Trial Tr. 1158 (Liggero); Depo. Desig. Tab 31 (02/13/07 Liggero) at 56. As Dr. Liggero further conceded, it is therefore unknown whether a higher speed could have been achieved had the sensitizing dye been added in conjunction with chemical sensitization. Trial Tr. 1158 (Liggero); Depo. Desig. Tab 31 (02/13/07 Liggero) at 63. In addition, Tani III discloses no information regarding the thickness of the T-grains described therein. Trial Tr. 1158 — 59 (Liggero); Depo. Desig. Tab 31 (02/13/07 Liggero) at 87-88.

Agfa argues that Tani V [T. Tani, "Photographic Sensitizing and Desensitizing Phenomena of Various Silver Halide Grains by Electron Traps," Photographic Science Eng'g, Vol. 17, No. 3, at 306-314 (May/June 1973) ("Tani V")] describes the optimal chemical sensitization of the tabular grain emulsions — "the degree of the increase of photographic speed due to these chemical sensitizations was compared for the various grains at their optimum conditions." TX 1031 (page 307), 1336; Trial Tr. at 1121 (Liggero).

Tani V does not disclose spectral sensitization of T-grains, optimal or otherwise. Depo. Desig. Tab 31 (02/13/07 Liggero) at 46 ("[T]his does not address spectral sensitization."); Trial Tr. 1157 (Liggero). In addition, Tani V discloses no information regarding the thickness of the T-grains described therein. Trial Tr. 1157-58 (Liggero).

Agfa argues that Tani VII [In T. Tani, "Photographic Effects of Electron and Positive Hole Traps in Silver Halides VII: Interaction Between Reduction Sensitization and Desensitization by Dyes on Various Silver Bromide Grains," Photographic Science Eng'g, Vol. 18, No. 5 (Sept/Oct. 1974) (TX 1033) ("Tani VII")] describes the chemical ("reduction") sensitization and the spectral sensitization of the same tabular grain emulsion. TX 1033, TX 1336; Trial Tr. at 1121-22 (Liggero). Tani VII discloses conditions sufficient for optimal chemical sensitization of the tabular grain emulsion — "50°C for 60 minutes." TX 1033 (page 570), 1336; Trial Tr. at 1121-22 (Liggero). Tani VII allegedly discloses conditions sufficient for optimal spectral sensitization of the tabular grain emulsion — "surface coverage of grains by dyes were relatively high." TX 1033 (page 570), 1336; Trial Tr. at 1121-22 (Liggero). According to Agfa, a person of ordinary skill in the art would have understood Tani VII to have disclosed optimally chemically and spectrally sensitized tabular grain films. Trial Tr. at 1121-22 (Liggero).

Tani VII relates to desensitization, and does not disclose optimal spectral sensitization or optimal chemical sensitization. Trial Tr. 1159 (Liggero); Depo. Desig. Tab 31 (02/13/07 Liggero) at 91. Dr. Liggero, Defendants' expert, stated on cross-examination that Tani VII does not explicitly disclose optimal spectral sensitization or optimal chemical sensitization. In addition, Tani VII discloses no information regarding the thickness of the T-grains described therein. Trial Tr. 1159 (Liggero); Depo. Desig. Tab 31 (02/13/07 Liggero) at 88.

Kodak filed U.S. Patent Application No. 06/320,899 ("the `899 application") on November 1, 1981, i.e., before the effective filing dates of any of the T-grain patents, naming as inventors Herbert Wilgus and Jong-Shinn Wey. TX 1015. On September 30, 1982, Kodak filed a continuation-in-part of the `899 application which resulted in U.S. Patent No. 4,414,306 to Wey and Wilgus. TX 1014.

The `899 patent application is admitted prior art against the `426 patent. TX 1006 (at EK00191) ("[the `899 patent application constitutes] a prior teaching of tabular grain silver chlorobromide emulsions in which the molar ratio of chloride to bromide ranges up to 2:3"). Further, the applications have different inventive entities.

The `899 application describes less than 0.2 tabular grains:

In a preferred form the emulsions are of high aspect ratio. As applied to the emulsions of the present invention the term "high aspect ratio" is herein defined as requiring that tabular silver halide grain containing chloride and bromide in at least annular grain regions having a thickness of less than 0.3 micron and a diameter of at least 0.6 micron have an average aspect ratio of greater than 8:1 and account for at least 35 percent of the total projected area of the silver halide grains.

Although emulsions according to the present invention can have average aspect ratios as low as 5:1, it is preferred that the emulsions have high average aspect ratios of greater than 8:1. Average aspect ratios can range up to 15:1, 30:1, or even higher. The preferred emulsions of the present invention have an average thickness less than 0.2 micron. In a preferred form of the invention these tabular grains account for at least 50 percent and optimally at least 70 percent of the total projected area of the silver halide grains containing chloride and bromide in at least annular grain regions.

TX 1015 (page 10). See also TX 1014 (`306 patent, col. 6, l. 52, to col. 7, l. 6).

The `899 application describes the optimal sensitization of the tabular grains — "the emulsions of the present invention are preferably, in accordance with prevailing manufacturing practices, optimally chemically and spectrally sensitized." TX 1015 (page 36). See also TX 1014 (`306 patent, col. 20, ll. 32-36). The `899 application also describes the use of the preferred tabular grain emulsions in an X-ray film format — "[w]hen the photographic elements of the invention are intended to serve radiographic applications. . . ." TX 1015 (page 46, EK107228). See also TX 1014 (`306 patent, col. 25, ll. 38-40).

Thus, Agfa concludes that the `899 application describes the asserted claims of the `426 patent — X-ray films comprising optimally spectrally sensitized, intermediate aspect ratio tabular grain emulsions. TX 1334 (¶¶ 39-43).

Kodak responds that the `899 patent application does not disclose a duplitized radiographic element containing an emulsion wherein tabular silver halide grains having a thickness of less than 0.2 micron and an average aspect ratio of from 5:1 to 8:1 account for at least 50 percent of the total projected area. In particular, the cited passage of the `899 application does not state that the average aspect ratio of grains less than 0.2µ is 5:1 to 8:1 as required by the `426 patent's claims. Nor does the passage indicate that those tabular grains having a thickness less than 0.2µ account for at least 50% of the total projected area.

A duplitized radiographic element was known in the art, and the `899 application referred to X-ray film format. The `899 application was cited by Kodak on a PTO 1449 form, and the examiner initialed the form next to the entry (as well as crossing out the entry, as was standard practice for patent applications listed on 1449 forms). It also is discussed in the T-grain patents. Thus, the `899 application was considered by the PTO and the claims allowed over the reference.

Kodak filed U.S. Patent Application No. 06/320,904 ("the `904 application") on November 1, 1981, naming as inventors Kofron, Booms, Haefner, Wilgus and Jones. TX 1008. On September 30, 1982, Kodak filed a continuation-in-part of the `904 application which resulted in U.S. Patent No. 4,439,520 to the same named inventors. TX 1003.

Agfa argues that that the `904 application renders the `426 patent claims invalid. Agfa bases its argument on Kodak's accusation that certain Agfa products infringe both the `520 and `426 patents under the doctrine of equivalents. See, e.g., TX 1055. Agfa argues that because Kodak alleges these Agfa products satisfy the high aspect ratio grain size requirements of the `520 patent and the intermediate aspect ratio grain requirements of the `426 patent, the intermediate aspect grain ratio of the `426 patent must be obvious in view of the high aspect ratio grain requirements of the `904 application and of the `520 patent. According to Agfa, the `904 application describes the use of the preferred tabular grain emulsions in an X-ray film format — "[w]hen the photographic elements of the invention are intended to serve radiographic applications. . . ." TX 312 (`520 patent, col. 41, ll. 31-36), 1008. The `904 application incorporates by reference Research Disclosure 18431 [TX 1036], which describes radiographic elements and crossover reduction. TX 312 (`520 patent, col. 41, ll. 31-36), 1008(71). Thus, it is Agfa's position that the `904 application describes the asserted claims of the `426 patent — X-ray films comprising optimally spectrally sensitized, intermediate aspect ratio tabular grain emulsions — if the `426 patent claims are permitted to cover the same subject matter claimed by the `520 patent. TX 1334 (¶¶ 35-38).

The questions to be resolved involve the scope of the claims and the impact of the prior art, not Kodak's arguments regarding equivalency. It is the properly interpreted claim that is reviewed, not a patent owner's accusation of infringement.

The `904 application does not disclose a duplitized radiographic element containing an emulsion wherein "tabular silver halide grains having at thickness of less than 0.2 micron and an average aspect ratio of from 5:1 to 8:1 account for at least 50 percent of the total projected area" as recited in claim 1 of the `426 patent.

As the party asserting the defense of invalidity, Defendants were required to establish invalidity as to each claim by clear and convincing evidence. In the present instance, Defendants have shown that pieces of claims were known in the art. What Defendants have not done, at least not clearly or convincingly, is establish that each asserted claim was either anticipated or obvious. Rather, Defendants have merely cited to various items of prior art and frequently seem to draw unsupported conclusions based upon that art. Defendants did not clearly and convincingly cite to prior art and apply it to the asserted claims in a manner sufficient to establish claim invalidity. Furthermore, to the extent Defendants rely upon obviousness, they have not adequately addressed the "secondary considerations" that apply to the obviousness determination.

An assessment of obviousness requires consideration of objective factors potentially indicating non-obviousness. Graham v. John Deere Co., 383 U.S. 1, 17-18, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966). "Such secondary considerations as commercial success, long felt but unsolved needs, failure of others, etc., might be utilized to give light to the circumstances surrounding the origin of the subject matter sought to be patented. As indicia of obviousness or nonobviousness, these inquiries may have relevancy." Graham, 383 U.S. at 17-18, 86 S.Ct. 684. Secondary considerations including a finding of commercial success may not carry sufficient weight to override a determination of obviousness based on primary considerations. Such a determination can be made as long as the secondary considerations are contemplated by the court. Ryko Mfg. Co. v. Nu-Star, Inc., 950 F.2d 714, 719 (Fed. Cir. 1991).

Kodak's medical imaging products incorporating the patented T-grain technology were publicly acclaimed, with experts and commentators acknowledging the achievement of previously unobtainable advantages. See, e.g.,

TX 472 ( Kodacolor VR 1000 — A Review, The British Journal of Photography, dated November 26, 1982) at EK106593 ("In emulsions which have been the norm until the present development, faster films have been produced by, at the simplest, producing larger grains of silver halide, usually requiring thicker emulsion layers, using more silver and resulting in coarser grained images. By changing the shape of the crystals by flattening them they intercept more light for a given amount of silver halide and the emulsion layer thickness, important for sharpness, is kept down. This fundamental change in the shape and structure of the silver halide crystals, named T-grain' by the Kodak research workers as an abbreviation for tabular grain, is said to be the company's single biggest advance in silver halide emulsion technology in more than fifty years")

TX 429 ( Eastman Color High-Speed Negative Film 7292, SMPTE Journal, dated September 1986) at EK062811 ("In 1983, Eastman Kodak Co. demonstrated a significant advance in photographic emulsion technology with the introduction of Kodacolor VR 1000 film, which incorporates Kodak T-grain emulsions. This class of emulsions, characterized by their tablet-like crystal structure, provides previously unobtainable advantages")

TX 387 ( Screen Film Processing Systems for Medical Radiography: A Historical Review, Radiographics, dated November, 1989) at EK049010, EK049017 (stating that introduction in 1983 of a film for medical imaging that had tabular grains was the result of a "breakthrough in emulsion technology that produced flat, tabular silver halide crystals, oriented with the flat side parallel to the support," and identifying, among others, Dr. Russell Holland, E.I. DuPont de Nemours and Company, "for providing historical information and helpful discussions on the article")

Other articles lauded the T-grain technology as it applied to other films, such as color photographic film.

TX 475 ( Kodak — fastest ever print film, Amateur Photographer, dated October 16, 1982) at EK106604 (calling Kodak's Kodacolor VR 1000 T-grain film "a combination of existing technology with a radical approach to conventional image recording")

TX 471 ( Fast-Film Coup, A colorful marvel from Kodak, Time Magazine, dated October 18, 1982) ("The company's new film represents an important technological break-through. For years researchers seeking to devise high-speed color print films have been stymied by the difficulties involved in increasing the light sensitivity of photographic film without producing grainy or fuzzy pictures. Kodak scientists overcame this problem by in effect redesigning the physical structure of the silver halide crystals that form the light-sensitive coating of unexposed film. In their changed shape the crystals now are flatter, with more of their surface area being exposed to light on the film itself")

TX 473 ( Eastman Kodak unveils "world's fastest" color film, Chicago Tribune, dated October 8, 1982) ("But some analysts said they expect the new emulsion technology to be used eventually to improve sharpness of a wide range of amateur films and to be used in such industrial products as X-ray, graphic arts and motion picture films")

TX 473 ( Eastman Kodak unveils "world's fastest" color film, Chicago Tribune, dated October 8, 1982) (stating that "it was Kodak's announcement of a new film crystal breakthrough — which the firm called the `single biggest advance in silver-halide emulsion technology in more than 50 years' — that apparently impressed Wall Street analysts and investors. In trading on the New York Stock Exchange, Kodak closed up $2.75 a share to $85.75, and was the fifth most active stock, with 672,600 shares changing hands");

Furthermore, Drs. Boom and Kofron received special recognition for their work. TX 428 ( Kodak Employees Accept Progress Medal Awards, PSA Journal Official Magazine of the Photographic Society of America, dated November 1988) at EK062800 ("In recognition of their pioneering research in photographic science; [f]or their historic breakthrough in the practical application of tabular silver grain structures; [f]or their research in photographic emulsions which has resulted in outstanding improvements in sensitivity, image sharpness, and granularity; [t]he Photographic Society of America is proud to present its highest award, the PSA Progress Medal for 1988, to: Dr. Robert E. Booms and Dr. James T. Kofron"); Trial Tr. 66-67 (Booms). Drs. Booms and Kofron also received the Inventor of the Year Award from the Rochester Intellectual Property Law Association in 1991.

Kodak's X-ray films employing the T-grain technology have been commercially successful. See, e.g., Trial Tr. 66 (Booms) (Kodak commercialized X-ray films products, known as the TMAT line of films, under the T-grain patents and those products were commercially successful); Trial Tr. 366-67 (Dillenbeck) (TMAT G falls within the scope of the `425 patent); TX 632 (P L for HI Analog Film Dental Extra-Oral 1996-2004) (showing hundreds of millions of dollars in sales of Kodak X-ray films employing T-grain technology); Depo. Desig. Tab 35 (06/05/03 Roberts) at 221 (Kodak products employing the T-grain technology have been commercially successful).

Kodak's competitors obtained licenses from Kodak for the T-grain technology. See, e.g., TX 171 (acknowledging that Fuji had taken a license from Kodak for T-grain `426 patent technology); TX 426 (Cross-licensing agreement between Eastman Kodak Company and Fuji, dated 05/26/1988) (including T-grain patents); and TX 459 (Cross-licensing agreement between 3M Company and Eastman Kodak Company, dated 03/01/1989) (including `426 T-grain patent).

Finally, in the Reexamination of the `426 patent, the PTO specifically found that "[e]ven if a prima facie case exists, the patent owner has demonstrated an unexpected improvement in speed and crossover relationship for radiographic elements utilizing the T-grains of the present invention sufficient to contradict the prima facie case." TX 423 (at EK060572).

Defendants' attempts to discredit this collection of secondary evidence have been unpersuasive.

It was incumbent upon Defendants to make a prima facie showing of invalidity before Kodak need address the validity of its patents. Patents are presumed valid and Kodak need not establish the validity of its patents. While the citation of art not considered by the PTO might allow the clear and convincing standard to more easily be met, the burden is always upon the party asserting invalidity to prove it.

In view of the above, it is recommended that Defendants have failed to establish invalidity of any claim of the asserted patents.

Literal Infringement by the Accused Products

Kodak accuses the following products of infringing the T-grain patents as set forth in the following table.

Trade name `426 patent claims `425 patent claims `520 patent claims Cronex 10T Independent claim 1 and dependent claims 2, 3, 5, and 6 Cronex 10TL Independent claim 1 and dependent claims 2, 3, 5, and 6 Ultravision C Independent claim 1 and dependent claims 2, 3, 5, and 6 Ultravision Ci Independent claim 1 Independent claim 1 Independent claim 1 and dependent claims and dependent claims and dependent claims 2, 3, 5, and 6 2, 3, 5, 6, and 7 2, 4, 5, 8, 9, 13, 16, and 28 Ultravision L Independent claim 1 Independent claim 1 Independent claim 1 and dependent claims and dependent claims and dependent claims 2, 3, 5, and 6 2, 3, 5, 6, and 7 2, 4, 5, 8, 9, 13, 16, and 28 Orthovision G (aka Independent claim 1 Independent claim 1 Independent claim 1 Adox Green) and dependent claims and dependent claims and dependent claims 2, 3, 5, and 6 2, 3, 5, 6, and 7 2, 4, 5, 8, 9, 13, 16, and 28 Orthovision L Independent claim 1 and dependent claims 2, 3, 5, and 6 Agfa CP-BU (aka Independent claim 1 Independent claim 1 Independent claim 1 Agfa B Plus, DI and dependent claims and dependent claims and dependent claims Image Plus Blue, 2, 3, 4, 5, and 6 2, 3, 5, 6, and 7 2, 3, 4, 5, 8, 9, 10, 13, Radiomat B Plus, 16, and 28 McKesson Clear Choice Performance Plus High Speed Blue, and Agfa Cronex 5) Independent claim 19 Independent claim 19 and dependent claim 20 Curix Ortho HT-G Independent claim 1 (aka Radiomat SG and dependent claims and McKesson Clear 2, 3, 5, 6, 7, and 8 Choice Performance Plus Medium Speed Green) DI Image Plus Green Independent claim 1 (aka Radiomat G and dependent claims Plus) 2, 3, 5, and 6 Opthos H Independent claim 1 Independent claim 1 Independent claim 1 and dependent claims and dependent claims and dependent claims 2, 3, 5, 6, 7, and 8 2, 3, 5, 6, 7, 8 and 9 2, 3, 4, 5, 8, 10, 13, 16, and 28 Making, using, offering to sell, and selling a patented invention all constitute infringement if done without the authority of the patent holder. 35 U.S.C. § 271(a). The determination of patent infringement is a two-step analysis: (1) interpreting the meaning and scope of the asserted patent claims and (2) comparing the accused product to the claims as interpreted. Markman v. Westview Instruments, Inc., 52 F.3d 967, 976 (Fed. Cir. 1995) (en banc), aff'd, 517 U.S. 370, 116 S.Ct. 1384, 134 L.Ed.2d 577 (1996). Kodak has the burden of establishing infringement by a preponderance of the evidence. Centricut, LLC v. Esab Group, Inc., 390 F.3d 1361, 1367 (Fed. Cir. 2004).

The claim construction of the patents in suit was set forth in the Report and Recommendation for Construction of Group I Patents, March 21, 2007 ("Claim Construction Recommendation"). Based on that claim construction, there are factual issues relating primarily to sensitivity and grain size.

The term "chemically sensitized" was defined in the Claim Construction Recommendation as the treatment of grains with a chemical which is not selective with respect to a spectral region for increasing photographic speed of the treated grains.

There is no dispute that each of Agfa's accused products is chemically sensitized. Agfa Statement of Fact ("SOF") No. 155 ("In all instances, the Agfa films are chemically sensitized").

The term "spectrally sensitized" has been construed in the Claim Construction Recommendation to mean the treatment of grains with a spectral sensitizer which is selective with respect to a spectral region and which causes an increase in sensitivity to the grain in that spectral region.

There is no dispute that each of Agfa's accused green light sensitized films are spectrally sensitized. Agfa SOF No. 156 ("In all instances, the accused green light sensitized films are spectrally sensitized. These films are Orthovision G, Orthovision L, CXODGUV2, CXODGGV2, CX1X5 and CXODGAS.").

Agfa argues that the remaining accused products — i.e., Cronex 10T, Cronex 10TL, Ultravision G, Ultravision C, Ultravision Ci, and Ultravision L — are not spectrally sensitized. According to Agfa, the socalled "DF-629" or B6B dye used in production of these products is not a spectral sensitizing dye. TX 1182, 1189, 1341 (¶ 178-99); Trial Tr. at 857-58, 1061-62, 1090-94 (Hudson), 992-1000 (Van den Zegel).

The overwhelming evidence points to a conclusion that all of Agfa's accused products are spectrally sensitized. The DF-629/B6B dye meets the T-grain patents' definition of a spectral sensitizer. The dye is selective with respect to a spectral region and causes an increase in sensitivity of the grain in that spectral region. See Trial Tr. 262-63, 265-66 (Marchetti) ("the grains are much more sensitive after the dye is added to the blue region of the spectrum"); TX 1189 (at A631497) (sensitivity of emulsion to blue light ( i.e., speed) shown to be higher when B6B/DF-629 dye is used (test no. 34902705) versus when dye is not used (test no. 349502702)); Trial Tr. 1062 (Hudson) (TX 1189 demonstrates that if you "leave the dye out," then "it's a slower film"); Trial Tr. 1206 (Marchetti) (TX 1189 indicates that the dye "fits the Special Master's definition" in that "it increases the speed in a spectral region"); Depo. Desig. Tab 2 (01/29/04 Apple) at 92-93 (Cronex 10T is "spectrally sensitized" using a "blue absorbent dye"); Depo. Desig. Tab 17 (07/08/04 M. Guy) at 97 (Agfa Rule 30(b)(6) witness on product formulations) ("I know when you add the [B6B/DF-629] dye to the emulsion, you increase the speed of the emulsion."); TX 833 (08/24/06 Marchetti expert report) ¶¶ 20-30 (B6B/DF-629 dye is a spectral sensitizing dye).

The record is replete with admissions and internal DuPont/Sterling documents expressly identifying the B6B/DF-629 dye as a spectral sensitizing dye or identifying the products containing this dye as having been spectrally sensitized. TX 181 ("Sterling Diagnostic Imaging Brevard R D Technology Portfolio — April 6, 1999") at A163979 (DF-629 is a "blue J-aggregating spectral dye"); TX 24 (Sterling "Patent Clearance for Cronex(r) 10ST (Type 578)") (A006027-A006165) at A006131 (one "component" of the product is "spectral sensitization"); TX 157 ("Low Cost, Half-Speed Blue Medical Screen Film Candidate Recommendation March 17, 1997") at A160041 ("Blue Spectral Sensitization" listed as a "Product Feature"); TX 16 ("Update on Patent Protection for Ultraviolet Imaging System") at A004982 (referring to DuPont tabular grain medical X-ray films "which are spectrally sensitized to the blue or UV"); Trial Tr. 935 (Hudson) (the B6B dye was referred to in DuPont/Sterling as a spectral sensitizing dye), Depo. Desig. Tab 22 (10/12/04 Hudson) at 36, 58-59 (the Sterling emulsion is "subjected to chemical and spectral sensitization"; a feature of the film is "blue spectral sensitization").

DuPont obtained U.S. Patent No. 5,108, 887 (TX 905), which is directed to the B6B/DF-629 dye. See TX 181 ("Sterling Diagnostic Imaging Brevard R D Technology Portfolio — April 6, 1999") at A163979 ("DF-629 . . . is protected by U.S. 5,108,887"), TX 18 at A005083 ("The dyes claimed [in the `887 patent] are those used for all current blue and UV tabular grain products."). The title of this DuPont patent is "Zeromethine Merocyanine Dyes as J-Aggregating Spectral Sensitizers for Tabular Emulsions." TX 905 (emphasis supplied). The specification of the DuPont `887 patent states, among other things, that "J-aggregation causes a single dye, or a mixture of dyes, to shift the absorption maxima to a longer wavelength." TX 905 (at col. 1, lines 50-52).

That the B6B/DF-629 dye is a spectral sensitizing dye is also established by a wedge spectrogram generated in the DuPont laboratories and reproduced in an internal DuPont Technical Report authored by DuPont scientist Dr. Dietrich Fabricius in 1990. See TX 280 (Technical Report entitled "Manufacture of New Sensitizing Dye for Cronex(r) 10T: DF-629 and D-905 Solution," authored by Dr. Fabricius) at A617343 (showing wedge spectra for DF-629); id. at A617339 ("Careful inspection of the DF-629 spectra reveals a spike near 460 nm, which is attributed to the DF-629 J-aggregate."); TX 843 (original DuPont Laboratory Notebook containing the DF-629 wedge spectrogram) at A054232-A054237 (data "indicates significant spectral sensitization from these dyes and especially DF-629."). The spike shown in this wedge spectrogram would not occur if the dye were not acting as a spectral sensitizer. Trial Tr. 269, 307-08 (Marchetti) (wedge spectrogram made by DuPont "absolutely" indicates that the dye is a spectral sensitizer). Dr. Fabricius (who also was an inventor in the DuPont `887 patent), was an expert on sensitization of silver halide grains and, in fact, "was the leading guy in terms of sensitizing dyes" at DuPont/Sterling. Trial Tr. 1077 (Hudson)

Based on the testimony of Dr. Hudson, Agfa argues that the wedge spectrogram does not necessarily indicate a spike near 460 nm. Agfa's argument is contradicted by a DuPont document, TX 280 at A617339, which states that "[c]areful inspection of the D-629 spectra reveals a spike near 460 nm, which is attributed to the DF-629 J-aggregate." Agfa also asserts that "DuPont did not perform the wedge spectrogram analysis using an equal energy source and appears to have used a glass slide-both of which distort the density bands," but cites no support for these assertions or other evidence that, if true, these conditions would have meaningful consequences.

Agfa argues that its dyes act as "site directors" to enhance chemical sensitization without functioning as a spectral sensitizing dye:

Dyes used in this manner can be referred to as "site directors" because they direct chemical sensitizers to certain locations within the grain, thereby focusing the energy absorbed by the grains to these locations. . . . [Speed increase alone is not proof of] spectral sensitization because site direction also produces a speed increase as a result of increased chemical sensitization without the dye necessarily functioning [as a spectral sensitizing dye].

TX 1341 (¶ 184); see also Trial Tr. at 1061-62 (Hudson).

To support its position, Agfa conducted experiments in which emulsions were sensitized using the of sensitization compounds. In the experiments, the chemical sensitization process was conducted until the maximum speed (at an acceptable fog level) was obtained. With the exception of the experimental variables of dye amount and the actual moment of dye addition, no other parameters were changed other than to demonstrate that higher levels of dye addition would not increase speed. TX 1341 (¶ 192-195); Trial Tr. at 1063-66 (Hudson), 992-1000 (Van den Zegel).

In each experiment, a first test was conducted in which the dye was present during chemical sensitization. That system was assigned a reference speed of zero. In a second test, speed was measured with no dye present. For all products, there was a speed loss relative to the first test. It is Agfa's position that the speed loss is due to either less efficient chemical sensitization (due to no site directing effect of the dye) or no spectral sensitization, or both. To determine if there is a loss due to no spectral sensitization, a third test was conducted in which the dye was added after chemical sensitization. Agfa's theory is that because the dye is added after chemical sensitization, the dye cannot act as a site director for chemical sensitization in the third tests.

For all products other than the green light sensitizer films, the third tests produced a speed loss greater than the second test, which included no dye. Agfa concludes that the dye does not function as a spectral sensitizer, but as a site director to enhance chemical sensitization. This conclusion is based on Agfa's theory that the dye would always increase sensitivity, regardless of whether it is added before or after chemical sensitization, if the dye was functioning as a spectral sensitizer. TX 1341 (¶ 191). Because the dye for these products only increased speed if it was added prior to the chemical sensitizer, Agfa concludes that the dye must function exclusively as a site director.

Agfa's tests are filled with too many unanswered questions and doubts to overcome the overwhelming evidence and admissions that the dyes are spectral sensitizers. Further, there are reasonable explanations to reconcile Agfa's test results with evidence that the dyes are spectral sensitizers. First, even if the B6B/DF-629 dye acts as a site director in the accused products, it also may be functioning as a spectral sensitizer. Trial Tr. 304-06, 320-22 (Marchetti) ("A dye cannot function only as a site director because if it does, the system will lose speed because the light that is absorbed will be lost."), 323 (Marchetti) ("[If] [the B6B and B6E/DF-629 dyes were] acting only as site directors, there may be some small increase in speed, but there would be a huge decrease in speed due to the fact that you're wasting all the light.").

Second, Dr. Van Den Zegel, who performed the tests, testified in his deposition that, when optimally sensitizing an emulsion, "[t]he amount of spectral sensitizer that is applied before chemical sensitization is greater than the amount that would be added if the spectral sensitizer would be added after chemical sensitization" and that where the spectral sensitizing dye is added after chemical sensitization, "too much spectral sensitizer gives desensitization." Depo. Desig. Tab 44 (05/26/04 Van Den Zegel) at 372-73. The experiment should have used a different dye level and/or varied other conditions of sensitization when the dye was added after chemical sensitization. Trial Tr. 1206-07, 1226 (Marchetti); Depo. Desig. Tab 31 (02/13/07 Liggero) at 267-68 (admitting that it is unfair to compare optimized emulsions with sub-optimized emulsions). The experiments conducted by Dr. Van den Zegel appear to show only that too much dye was used, resulting in desensitization, in the case where the dye was added after chemical sensitization.

Accordingly, it is my recommendation that all of the accused products be found to be spectrally sensitized.

The phrase "substantially optimally chemically and spectrally sensitized" has been defined in the Claim Construction Recommendation as emulsions which have been both chemically and spectrally sensitized to "achieve speeds of at least 60 percent of the maximum log speed attainable from the grains in the spectral region of sensitization under the contemplated conditions of use and processing." Log speed is 100 (1-log E), wherein E is measured in meter-candle seconds at a density of 0.1 above fog.

The phrase "substantially optimally chemically sensitized" has been defined in the Claim Construction Recommendation as emulsions which have been chemically sensitized to "achieve speeds of at least 60 percent of the maximum log speed attainable from the grains in the native (blue) spectral region under the contemplated conditions of use and processing." Log speed is 100 (1-log E), wherein E is measured in meter-candle seconds at a density of 0.1 above fog.

These definitions incorporate within them the concept that optimization is conducted taking into account the contemplated conditions of use and processing in the photographic arts. Dr. Marchetti, Kodak's expert, agreed with these definitions. Trial Tr. 282. The evidence of record is clear that persons of ordinary skill in the photographic arts would have interpreted optimum sensitization under contemplated conditions of use and processing to mean the maximum log speed attainable from the grains while meeting the other product specifications, such as may be specified by the photographer or user. These other product requirements may include fog levels, contrast, and keeping properties. Trial Tr. 62 (Booms) ("[O]ptimal sensitization means that you get the best sensitivity that you can while meeting the other system requirements of whatever product that you're working on."); Trial Tr. 278-80 (Marchetti) ("Optimal sensitization is essentially a practical definition to optimize the emulsion to get the most [speed] that you can within the constraints that you are going to use it for."); Trial Tr. 1104 (Liggero). "Optimal sensitization" as that term is commonly used in the industry and in the T-grain patents does not mean obtaining maximum photographic speed in an absolute sense. Trial Tr. 279 (Marchetti) (persons skilled in the art do not understand optimal sensitization to mean the absolute highest speed without regard to any other characteristics), 280 (Marchetti) (not aware of anyone in the photographic community that sensitizes without regard to other product requirements), 279 (Marchetti) ("[I]f you don't [take into account the other system requirements] you don't produce a product that is useful for anything or a product that meets the customer demands. . . ."), 937-38 (Hudson); Depo. Desig. Tab 32 (02/14/07 Liggero) at 334 (does not know of any company "whose manufacturing practice is to try to reach a maximum speed without regard to any fact — other factors such as fog").

Kodak never undertook any experiments to quantify the extent of sensitization of the accused products. However, a patentee may prove direct infringement by either direct or circumstantial evidence. Moleculon Research Corp. v. CBS, Inc., 793 F.2d 1261, 1272 (Fed. Cir. 1986), cert. denied, 479 U.S. 1030, 107 S.Ct. 875, 93 L.Ed.2d 829 (1987), rev'd on other grounds, 872 F.2d 407 (Fed. Cir. 1989). There is no requirement that direct evidence be introduced, nor is a jury's preference for circumstantial evidence over direct evidence unreasonable per se. See Fuji Photo Film, Co. v. Jazz Photo Corp., 394 F.3d 1368, 1374 (Fed. Cir. 2005); Moleculon Research, 793 F.2d at 1272 (noting "it is hornbook law that direct evidence of a fact is not necessary"); Liquid Dynamics Corp. v. Vaughan Co., 449 F.3d 1209 (Fed. Cir. 2006), cert. denied, ___ U.S. ___, 127 S.Ct. 599, 166 L.Ed.2d 432 (2006) ("Though it is [patentee's] burden to prove infringement, it chose to rely on circumstantial evidence for some installations. While direct evidence may have made [patentee's] case of infringement stronger, it does not render the case presented to the jury unreasonable as a matter of law. Indeed, the jury could have reasonably inferred from the evidence submitted that each of the tanks infringed the claims"); see also Michalic v. Cleveland Tankers, Inc., 364 U.S. 325, 330, 81 S.Ct. 6, 5 L.Ed.2d 20 (1960) ("Circumstantial evidence is not only sufficient, but may also be more certain, satisfying and persuasive than direct evidence.").

There is at least a preponderance of evidence sufficient to find that the accused products meet the "substantially optimally" sensitized claim limitations. DuPont, Sterling and Agfa employed designed experiments to sensitize the accused products to achieve the maximum speed consistent with other system requirements. The DuPont/Sterling and Agfa scientists who conducted such experiments were skilled in that art. Consistent with standard industry practice, Agfa acknowledged that the goal of the designed experiments was to achieve the highest speed taking into account the contemplated conditions of use and processing. Trial Tr. 1088 (Hudson) ("[W]e were almost always trying to strive to increase to get more out of the products that we could."), 284-88, 298 (Marchetti) (describing evidence regarding designed experiments conducted by Defendants); see also Depo. Desig. Tab 22 (10/12/04 Hudson) at 57-60 (it was the "general goal" on all of the Sterling tabular grain X-ray films to get the best speed consistent with the fog and fog-aging parameters that were applicable); id. at 94 (referring to use of designed experiments to "try to find the optimum levels"); Depo. Desig. Tab 23 (02/15/07 Hudson) at 163 (scientists who conducted such experiments were skilled in the art).

At trial, Agfa's expert Dr. Hudson conceded that the accused products were developed using designed experiments with the intent of maximizing speed while simultaneously satisfying other product specifications.

Q. Now, Dr. Hudson, it's true, is it not, that the people that developed products at DuPont and Sterling tried to get all the speed they could out of emulsions while maintaining acceptable fog and fog aging levels?

A. Yes. And then there is [sic] also other [parameters] as well, but that's right.

Q. And the goal was to get the best speeds possible with the fog and fog aging parameters applicable, isn't that correct?

A. Yes. But you still have to have acceptable DMAX, contrast, safelight, for the customer, physical properties that the film itself, and, of course, the fog aging, that is the normal shelf life, just so we're clear.

Q. The people at DuPont and Sterling conducted designed experiments to find the optimal level for those products, correct?

A. We conducted designed experiments and tried to get find the best speed we could within the constraints of having, you know, our aging test, acceptable fog contrast, these [sic] hardness of the film samples. And then that's why we were able to find things that led us to go from 10TI and 10ST and make the big grain drop.

Q. And you used designed experiments in that process?

A. Yes, for that particular example we probably used 100 or 200.

Q. And those people that performed the designed experiments were skilled in the art?

A. Yes.

Trial Tr. 935-36 (Hudson).

Additional evidence concerning the optimal sensitization of Cronex 10T (also internally referred to as "10ST") and Cronex 10TL can be found at:

TX 157 ("Low Cost, Half-Speed Blue Medical Screen Film Candidate Recommendation March 17, 1997") at A160040 ("the methods for sensitizing these grains were optimized in the laboratory")

TX 181 ("Sterling Diagnostic Imaging Brevard R D Technology Portfolio — April 6, 1999") at A163983 (Sterling X-ray emulsions made use of "smaller/thinner tabular grains in which spectral sensitization has been fully utilized")

TX 162 ("New Cronex(r)10ST Medical Screen Film Technology Assessment Report") at A160788 ("The sensitizing dye and AF-165 have been increased to optimally sensitize" the grains)

TX 162 ("New Cronex(r)10ST Medical Screen Film Technology Assessment Report") at A160791 ("The elements of the `improved sensitization scheme' which are used in Cronex(r)10ST consist of optimized amounts of the dyes and unconventional addition times during digestion")

TX 24 ("97-B-MR-02 10ST Cronex(r) 10T, (Type 578) Consumables Manufacturing Recommendation") at A006032 (the dye, gold, mercury, TCA, and GDBAS used in the product "are all sensitizers whose amounts have been optimized")

TX 66 ("Product Design Plan Cover-sheet — Low Coating Weight 10TL (579)") at A027034 ("Aim speed . . . will be achieved by employing the same sensitization techniques which were successfully implemented into Cronex 10ST")

Evidence concerning the optimal sensitization of Agfa CP-BU (aka Agfa B Plus, DI Image Plus Blue, Radiomat B Plus, McKesson Clear Choice Performance Plus High Speed Blue, and Agfa Cronex 5) (internal designator CXIX5), can be found at:

TX 42 ("Product Development Curix Blue Mat (screen and film)") at A017054 ("The concentration of B6B that is added is critical and 150 ml/1000 ZZ is optimal for [the emulsions used in CX1X5]. Higher concentrations produce desensitization, lower concentrations produce doka.")

Depo. Desig. Tab 50 (06/11/04 Verbeeck) (Agfa scientist) at 95-98 (the emulsions in CX1X5 are optimally sensitized in the sense of "total sensitization," including "spectral and chemical"; the use of the spectral sensitizer increases the speed of the emulsion; if more spectral sensitizer were used, desensitization would occur; the amount of spectral sensitizer is sufficient to achieve total optimal sensitization)

TX 37 ("Components for a New Generation Curix UV-Blue Material" dated 10/07/96) at A015396 (table showing "Optimization of ripening solutions CX BLUE DG"; solutions include the B6B dye; "[t]his optimized ripening formula was brought into production") Id. at A015397.

Depo. Desig. Tab 54 (09/08/04 Ver-beeck) (Agfa scientist) at 495-96 (optimalization of the relationship between conservation, fog, speed, and doka was achieved in CX1X5)

Depo. Desig. Tab 33 (09/01/04 Mestdagh) (Agfa scientist) at 39-41 (with respect to the emulsions in Agfa-Gevaert Depo. Exh. 19 [TX 872; includes CX1X5], Agfa achieved at least 60% of the maximum speed achievable within the contemplated parameters reflecting the contemplated conditions and exposure; the dye was adsorbed to the surface of the tabular grains in an amount sufficient to optimally sensitize the tabular grains)

Additional evidence concerning the optimal sensitization of Curix Ortho HT-G (aka Radiomat SG and McKesson Clear Choice Performance Plus Medium Speed Green) (internal designator CXODGU.V2), DI Image Plus Green (aka Radiomat G Plus) (internal designator CXODGGV2), and Opthos H (internal designator CXOD-GAS) can be found at:

Depo. Desig. Tab 45 (05/27/04 Van Den Zegel) at 419-20 (the CXODGU.V2 and CXODGGV2 emulsions have been optimized in such a way that optimal properties are derived from the materials; that result is achieved in part by using a spectral sensitizer)

TX 57 ("Product Description Curix Ortho DGAS") at A019312 (AT006004) ("Co-ripening is optimal to a maximum sensitivity, with fog and doka [dark room sensitivity] being within the norm. . . .")

Depo. Desig. Tab 52 (06/15/04 Verbeeck) (Agfa scientist) at 254-55 (the emulsions in CXODGAS are optimally sensitized in the sense of "total sensitization" (spectral and chemical); "we have sought an optimal performance of the emulsion")

Depo. Desig. Tab 33 (09/01/04 Mestdagh) (Agfa scientist) at 39-41 (with respect to the emulsions in Agfa-Gevaert Depo. Exh. 19 [TX 872; includes CXOD-GUV2, CXODGGV2, and CXODGAS], Agfa achieved at least 60% of the maximum speed achievable within the contemplated parameters reflecting the contemplated conditions and exposure; the dye was adsorbed to the surface of the tabular grains in an amount sufficient to optimally sensitize the tabular grains)

While Kodak did not produce direct evidence of what the speed or maximum speed was for each of the Agfa products, the evidence points to a conclusion that all of Agfa's products were substantially optimally sensitized for their contemplated conditions of use and processing. The claim construction did not require that the maximum speed of the film be determined in order to conclude whether a given emulsion was optimally sensitized. Based upon statements in the patents, the claim construction articulated clearly that the contemplated conditions of use and processing needed to be taken into account and neither party filed an objection to the Claim Construction Recommendation. Indeed, at trial Kodak's expert Dr. Marchetti acknowledged that he agreed with the construction of that limitation. Trial Tr. 282. Accordingly, it is my recommendation that Agfa's products are all spectrally sensitized, substantially optimally chemically and spectrally sensitized, and substantially chemically sensitized.

Claim 1 of the `426 patent recites that the tabular silver halide grains having a thickness of less than 0.2 micron and an average aspect ratio of from 5:1 to 8:1 account for at least 50 percent of the total projected area of the silver halide grains present in the silver halide emulsion. Emulsions meeting these criteria are referred to as "intermediate aspect ratio emulsions."

Claim 1 of the `425 patent and claim 1 of the `520 patent recite that the tabular silver halide grains having a thickness of less than 0.3 micron, a diameter of at least 0.6 micron, and an average aspect ratio of greater than 8:1 account for at least 50 percent of the total projected area of the silver halide grains present in the silver halide emulsion. Emulsions satisfying these criteria are referred to as "high aspect ratio emulsions."

Claim 19 of the `425 patent and claim 19 of the `520 patent characterize the tabular silver grains as having a thickness of less than 0.5 micron, a diameter of at least 0.6 micron, and an average aspect ratio greater than 8:1 accounting for at least 50 percent of the total projected area of the silver halide grains present in the silver halide emulsion.

Kodak's TEM images were taken at magnifications between 8469x and 10,000x, TX 1341(135), whereas Agfa uses a higher magnification of 14,000x to 18,000x to measure the grain replicas. TX 1341 (¶ 20). Agfa alleges that Kodak's use of lower magnification levels resulted in measurement inaccuracies with respect to the projected surface area of the silver halide grain replicas, which in turn resulted in inaccuracies in the calculated aspect ratio and in the calculated percent projected area. TX 1341 (¶ 35).

An opposite case could be made, i.e., that Kodak's measurements are arguably more precise than those of Agfa's inasmuch as Kodak used a measuring device with increments of 0.01 millimeters, whereas Agfa used a measuring device with increments of 0.5 millimeters. Trial Tr. 1082 (Hudson). In any event, neither argument is persuasive. Dr. Hudson explained during trial how the testimony of Agfa's expert Dr. Apley demonstrated that any error in measurement caused by differing magnifications "averages out," assuming that at least 100 measurements are taken.

And the other thing that we just heard from Dr. Apley is because all these measurements are random, for the most part sometimes you're a plus, a hundredth of millimeter too long and on the next grain measuring you may get a hundredth of a millimeter too short. The beauty is all the measurement, the errors measure out he included the two sources of error. One part of was the measurement error; the other part was sampling error. The sampling error totally swamped the measuring error, so the point is we can knit pick all we want on the measurements, whether this one is more accurate than that. Once you do 100 of the measurements, the beauty is that error averages out.

Trial Tr. at 1083 (Hudson). Moreover, Dr. Cole, a former Kodak employee and expert with extensive experience in grain measurement, testified that the Kodak methodology used by Mr. Van Dam was accurate and had been implemented properly. Trial Tr. 188-92.

For these reasons, I find that no persuasive showing has been made that either Kodak's lower magnification or Agfa's greater measuring increments materially affects the accuracy of grain measurements by either of the parties, at least for those samples for which at least 100 measurements were taken.

Agfa asserts that Kodak, and in particular Mr. Van Dam, changed the procedures by which shadowed electron micrographs were obtained for the purposes of this action. In particular, Agfa asserts that shadowed electron micrographs taken for the purposes of this action were obtained with higher shadow angles, i.e., 21 to 33 degrees, than Kodak routinely used before this action. Trial Tr. at 223-33 (Cole). It is Agfa's contention that Kodak's shadow angles are biased toward underestimating the actual grain thicknesses and overestimating the percent total projected areas and average aspect ratios of the silver halide grains of the accused products. TX 1341 (¶¶ 37-38); Trial Tr. at 235 (Cole).

Agfa's assertion that Kodak has only recently used a higher shadowing angle is contradicted by the evidence. The materials provided by Kodak to DuPont during their 1995 meeting indicated that shadowing angles ranging from 15° to 35° may be used. TX 468 (at EK106311).

Moreover, the record reflects a reasonable explanation for selecting higher shadow angles. An approximate 18° shadowing angle is convenient because it results in a roughly 3:1 ratio between the length of the shadow and the height (thickness) of the grain, enabling quick thickness calculations. Trial Tr. 190 (Cole). A higher shadowing angle can provide advantages, in that it can result in fewer shadows falling on other structures in the TEM field, thereby producing more measurable grains within a given microphotograph. Trial Tr. 166, 168 (Van Dam) (a higher angle will produce a shorter shadow, making it less likely that a given shadow will impinge on a grain near it); Depo. Desig. Tab 32 (02/14/07 Liggero) at 350-51 ("Q. So is it correct to say that if you . . . heighten the shadow angle, you are going to have shadows falling on fewer structures? A. That's correct."). The shorter shadows produced by the higher shadowing angle can also be more crisp and clearly delineated. Depo. Desig. Tab 24 (02/16/07 Hudson) at 367 ("Q. But you will see greater gradient differences at the end of the shadows, correct? A. Yes, the contrast will he higher.").

Agfa's assertion that Kodak's high shadow angles produced inaccuracies relies principally on the trial testimony and expert report of Dr. Hudson. TX 1341. Kodak has rebutted this testimony with its own testimony that the shadowing angle used does not affect the accuracy of the grain thickness determination, because the shadow angle is taken into consideration in the formula for calculating the thickness. Trial Tr. 156 (Van Dam) (it is not true that a higher shadowing angle is more likely to result in understatement of grain thickness), 168-69 (Cole) ("the shadow angle itself is not an important parameter"), 189-90 (Cole) (there is a very large tolerance in the angle that can be used). Kodak also offered testimony that that the shadow length is immaterial because the formula for measuring thickness takes into account the shadow angle via measurement of the reference bead's diameter and shadow length. Trial Tr. 169 (Van Dam) ("I don't believe the actual shadow angle is important. It's determining the shadow angle."), 248-51 (Cole) (the holder from which the heavy metal is sputtered is not fixed or at a pre-set angle; the angle used does not affect the measurement), 217 (Cole) (key is determining "what the angle actually is"). Therefore I conclude that the shadow angles employed by Kodak did not lead to systematic distortion of the grain measurement data.

Even if the shadow angle raises an inaccuracy, Agfa did not demonstrate the magnitude of that inaccuracy or that the inaccuracy is meaningful. Depo. Desig. Tab 24 (02/16/07 Hudson) at 364 ("Q. And what is — in your — according to you, what is the amount of the inaccuracy resulting from the shadowing angles that Kodak used? A. I have not done an analysis to put a quantitative number on the error that would be introduced."); Depo. Desig. Tab 32 (02/14/07 Liggero) at 347 ("Q. Have you attempted to quantify the amount, if any, of any inaccuracy resulting from Kodak's use of the shadowing angles it uses? A. Let's see. Not that I recall."); see also id. at 352-53 (not aware of anyone having attempted, in any fashion, to quantify any "understatement" of grain thickness resulting from the shadow angle used by Kodak). As demonstrated by Dr. Cole's analysis, assuming rounded grain edges, the difference (in terms of the ultimate calculated thickness of the grain) between using a 25.5° shadowing angle versus an 18.1° shadowing angle is immaterial. TX 672 (Cole analysis); Trial Tr. 234-35 (Cole). Thus, the underestimation, if it should exist, may very well be negligible.

For these reasons, I find that no persuasive showing has been made that the shadow angle selected by Kodak resulted in any material inaccuracies in Kodak's grain measurements.

In the following diagram of a tabular grain replica, h_s represents the shadow length of a grain.

Exhibit

In measuring the projected area of each grain, Agfa includes the "dark margin" (on the right side of the diagram) and excludes the "white margin" (on the left side of the diagram). Agfa's stated rationale for including one of the margins but not the other in its measurements is that the TEM image of the carbon replica does not exactly represent the size of the silver halide grain. The gold and carbon layers formed in the TEM imaging protocol contribute an additional length to the TEM image at the margins of the grain. One edge of the silver halide grain is located somewhere in the dark edge in the TEM image and the opposite edge of the silver halide grain is located somewhere in the white edge in the TEM image. Agfa approximates the edges of the silver halide grain as extending half way into the black edge and half way into the white edge of the TEM image. According to Agfa, because the width of the dark edge is approximately equal to the width of the white edge in the TEM micrographs, a reasonable approximation of the length of the silver halide grain is obtained by including all of the dark edge and excluding all of the white edge.

Kodak describes Agfa's rationale as "spurious," asserting that any portion of the dark margin that extends beyond the actual boundary of the grain is negligible and beyond the resolution of Agfa's measuring device. Agfa's omission of the white margin in measuring h_s, according to Kodak, results in understatement of the projected area and, consequently, understatement of the equivalent circular diameter or ECD and aspect ratio. TX 836 (Exh. A), 831 (¶ 30). These alleged defects have not restrained Kodak from relying on Agfagenerated data to support its infringement case, however.

Kodak, on the other hand, used an electronic digitizer to measure the projected area of each grain, including within its measurement both the white margin and the dark margin. Agfa argues that Kodak's digitizer method overestimates the projected area of the silver halide grains (and consequently overstates grain diameter and aspect ratio) because it does not take into account the thickness of the platinum palladium and carbon layers. TX 1341 (¶ 39), 1244. Despite these alleged defects, Agfa likewise has not refrained from relying on Kodak-generated data where helpful to its case.

Each party relies primarily on the opinions of their respective experts as to whether or not one or both margins should be included in the measurement of h_s. Neither parties' expert was particularly persuasive over the others'. Further, none of the experts was able to persuasively demonstrate or quantify the percentage of the dark and white margins that is attributable to the gold, platinum palladium, and carbon layers. See Depo. Desig. Tab 32 (Liggero) at 338 (has made no quantification of any such "overstatement"); Depo. Desig. Tab 24 (02/16/07 Hudson) at 362 ("Q. But do you have any particular magnitude of overstatement in mind? A. No, I do not.")

Accordingly, it is my recommendation that neither party has demonstrated that the other systematically and meaningfully overestimated or underestimated its measurements based on the inclusion or exclusion of the white margin.

Agfa's testimony regarding the control it exerted over its manufacturing process suggests some level of homogeneity across populations broader than a single specific emulsion. Trial Tr. 813 (Hudson) ("Once we had settled in on a recipe, if you will, for making the grains for a particular application, then that recipe becomes standardized in the manufacturing area. And . . . in our manufacturing area we have very tight controls on the dispensing into these recipe solutions."), 835 (Hudson) ("Q. . . . [H]ow much variation was there in the manufacturing process? A. Other than our start up period and then an occasional equipment failure, the control was very good."); 964-65 (Van den Zegel) ("We agreed our emulsions were very reproducible and further analysis of each emulsion with carbon replica was not necessary."); Depo. Desig. Tab 16 (02/06/04 J. Guy) at 62 (grain thickness in the manufactured product had a "very, very tight distribution").

Yet, Agfa argues that film products do not have a perfectly uniform distribution of grains throughout the entire film sheet due to the nature of the processes necessary to make and coat the grains onto a substrate. Coatings can vary in distributions of grain sizes and morphology over a single coating run as well as over several coating runs using nominally the same emulsions. Over several years, the coatings can vary. Variability of grain sizes and morphology can occur from coating to coating even though each of the coating runs has acceptable photographic properties. TX 1341 (¶ 45).

Agfa contends that the grain size properties and distribution of a particular sheet of film cannot be extrapolated to characterize the grain size properties of different coatings, let alone an entire product line manufactured over the course of years. TX 1341 (¶ 46). Film sheets are cut from coating productions, often on the order of 100,000 square meters to 1 million square meters of film per coating run. The TEM method employed samples approximately a hole-punch sized piece of a sheet of film (about 14 x 17 inches). Only a small amount of the grains are found in the hole-punch sized sample. Then, only a smaller number of the grains from this sample find their way to an actual photomicrograph. The actual TEM method studies a sample of a sample of a sample, etc. TX 1341 (¶ 46). According to Agfa, as a result of the product non-homogeneity and the deficiencies of the TEM method, an accurate assessment of the grain size distribution for a commercial product would require the assessment of thousands of grains from each production run with samples analyzed at various locations over the rolls of film produced.

Considering that an emulsion may be coated onto substrate in excess of 100,000 square meters and yet the sample is substantially much smaller, then it makes logical sense that there will be variability across that substrate for any number of reasons.⁶ Kodak would have one assume that such a small sample is representative of such a large area of coated material. If the sample were truly representative, then there would seemingly be no need to resort to equivalency as to some runs of a given product and literal infringement of others. Likewise, the variability in results, as elsewhere discussed, would not have been as large if the materials were as homogenous as Kodak suggests.

6 In its opening, Agfa asserted that the size differential was similar to the area of the United States versus the area of a postage stamp.

Agfa's expert, Dr. Apley, admitted in his expert report non-homogeneity across broad populations, which Dr. Apley designated as a fifth source of error, cannot be quantified based on the experimental data available. TX 130 (¶ 16); Trial Tr. 1043-44 (Apley). While the error caused by relatively small sizes and non-homogeneity may not be quantifiable, it also cannot be completely ignored or discounted.

Some DuPont- and Sterling-originated products, such as Cronex 10TL, contained mixtures of two different size populations of grains ( i.e., a fine grain component and a course grain component). Agfa asserts that this built-in non-homogeneity and stratification of different grain populations in products such as Cronex 10TL leads to sampling errors. TX 1341 (¶¶ 48-49).

The longstanding commonly employed practice (at both Kodak and Agfa) is to measure 100-150 grains in a sample. TX 910 (U.S. Patent No. 5,595,864 to Van Den Zegel et al., assigned to Agfa, at col. 8, lines 20-22, reporting patent examples based on shadowed TEM testing of at least 100 grains); Trial Tr. 130 (Van Dam) ("I typically measure between 100 and 150 grains."), 194 (Cole) ("In my training at Kodak, I was first introduced to the grain sizing techniques in 1981 and at that time it was a standard procedure to count somewhere in excess of 100 grains. Five years later when I started doing emulsion precipitation work, we used that same figure. So at Kodak, it was the de facto number that was used generally."); Depo. Desig. Tab 9 (06/25/04 Claes) (Agfa employee) at 61 ("As a standard, we measure 100 crystals."); Depo. Desig. Tab 33 (09/01/04 Mestdagh (Agfa employee) at 89); Agfa's April 27, 2007 Pre-Trial SOF No. 168 ("Usually, at least 100 grain replicas for each sample are measured.");

The parties' own actions suggest that an analysis of 100-150 grains yields a result that is meaningful. Trial Tr. 1003 (Van den Zegel) ("Q. And Agfa has long understood that a grain sample of at least 100 grains provides a statistically significant analysis, correct? A. For the Mortsel products, yes."), 196, 197-99 (Cole) (between 100 and 150 grains, the standard error approaches zero), 240 (Cole) (under central limit theorem, the mean of a sample from a random population will not shift relative to the mean of that population). It is understood by persons skilled in the art that an analysis based on 100-150 grains is sufficient to determine whether a sample falls within the scope of the T-grain patents. Depo. Desig. Tab 30 (03/10/05 Liggero) at 42-43 ("In my view, the total projected area is obtained by measuring, say, 100 individual grains, 100 to 150 individual grains. And the area of each of the grains, the surface area of each of the grains, is calculated. That then represents the total projected area of all the grains."); Depo. Desig. Tab 23 (02/15/07 Hudson) at 266-67 ("Q. Do you know of any company in the world that measures more than 100 to 150 grains using the TEM process in determining whether or not a product is infringing the Kodak T-grain patents? MR. LOWE: Object to form. BY THE WITNESS: A. No, I'm not aware of a company that measures more than 100 to 150 grains using the TEM method to determine whether they infringe or don't infringe or to characterize the emulsions according to those patents.").

Sources of error that arise in sampling and measuring grain sizes from Kodak's TEMs were assessed by Dr. Apley. In his report, Dr. Apley focused on two sources of error: (1) random sampling error in selecting the sample of crystals; and (2) random measurement error in the repeatability of measurements of the same crystals taken by the same operator.

To calculate the potential error in random sampling error and random measurement error, a bootstrap approach was used by Dr. Apley to establish a 95% joint confidence region. The 95% confidence region corresponds to plus or minus two standard deviations, which is common for statistical analysis. Trial Tr. 1045 (Apley). Error bars were then established to permit assessment of whether a film sample grain measurement analysis while nominally appearing to fall within the literal scope of the claimed grain size parameters actually may fall outside of the literal scope of those parameters. TX 1340 (¶¶ 20-26); Trial Tr. at 1028-32 (Apley).

It bears mentioning that Dr. Apley did not take into account in his calculations the following additional potential sources of error: (i) sampling bias due primarily to human subjectivity when choosing which crystals to measure ("Source 3" error); (ii) measurement bias due primarily to human subjectivity in how the measurements are taken ("Source 4" error); and (iii) nonho-rnogeneity across populations that are broader than a single specific emulsion ( e.g., a roll of film) ("Source 5" error). These sources of error could not be accurately quantified under the circumstances according to Dr. Apley. TX 1340 (¶¶ 14-16).

I find the testimony of Dr. Apley to be persuasive and largely unrebutted by Kodak. Dr. Apley's error analysis goes a long way towards explaining the variability of grain measurements taken by each of the parties. Indeed, Dr. Apley's analysis helps to explain why for given film products Kodak sometimes asserts literal infringement as to one sample but equivalency as to another.

Kodak criticizes use of the "bootstrap" analysis because the analysis is not mentioned in the T-grain patents. Trial Tr. 1049 (Apley); Depo. Desig. Tab 1 (01/04/07 Apley). However, the T-grain patents also do not expressly mention information pertaining to measurement protocols, such as suitable sample size, whether or not to include one or both grain margins of a shadowed electron micrograph in a measurement, shadowing angle, etc. As evident from the above discussion, clearly information such as sample size, while not mentioned in the T-grain patents, is relevant to the issue of infringement. Furthermore, Kodak has not attempted to establish that bootstrap analysis is unreliable, statistically unsound or otherwise unsuited for the task to which it was put. Dr. Apley's expert report includes citations to various literature sources establishing that the bootstrap methodology meets the requirements of Rule 702 of the Federal Rules of Evidence. Kodak did not object to the use of bootstrap analysis under Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993) as somehow being inappropriate for consideration.

Kodak argues that the "bootstrap" analysis is not commonly applied by persons skilled in the art to measurements from shadowed TEMs. Trial Tr. 1144 (Liggero); Depo. Desig. Tab 23 (02/15/07 Hudson) at 221. The same could be said for the use of the shadowed electron microscope technique for emulsion manufacturing, i.e., TEMs are not commonly applied by persons skilled in the art during emulsion manufacturing. During manufacturing, neither Kodak nor Agfa use TEMs to monitor the consistency of their products. The TEM method identified in the tabular grain patents to measure the grains is impractical on a commercial scale. TX 1341 (¶ 43); Trial Tr. at 964-65 (Van den Zegel).

Dr. Apley testified that the bootstrapping method is a common method employed by statisticians. Trial Tr. 1030 (Apley). His expert report and testimony concerning the effects of random sampling error (Source 1 error) and random measurement error (Source 2 error) of Kodak's data have gone largely unrebutted.

Kodak also criticizes the bootstrapping analysis on the ground that any measurement variability error in the grain sizing data generated by Kodak was insubstantial or de minimus. Kodak points to, inter alia, Dr. Apley's testimony referring to measurement variability as "insignificant". Dr. Apley qualified that "[t]he lions [sic] of the variability came from the sampling error," i.e., the first source of error considered in his report, not measurement variability. Trial Tr. 1044 (Apley). Further, the error range computed by Dr. Apley's bootstrapping technique cannot be dismissed as insignificant. As discussed infra, in several instances where Kodak alleges that a TEM analysis of an accused product sample finds a tabular grain distribution falling within the literal scope of an asserted claim of one of the tabular grain patents, that product sample may just as likely fall outside of the literal scope of the asserted claim when just sampling and measurement errors are considered as described and calculated by Dr. Apley. Further, several product batches which are purported to be homogeneous produced both infringing samples and non-infringing samples. Indeed, any physical measurement, no matter how precise, is subject to some level of error. In its post-trial Statement of Fact No. 112, Kodak acknowledged "grain measurements data produced in this case show that a number of accused products . . . sometimes fall within the definition of an Intermediate or High Aspect Ratio Emulsion and other times fall just outside those definitions".

Dr. Apley testified without any substantial rebuttal that taking the error into account as determined through the bootstrap methodology, one could not state with confidence where exactly a sampled data set fell. In other words, taking the error into account, one could not with confidence conclude that a given sample fell within or outside the literal scope of claims as explained below. All one could state, according to Dr. Apley, was that the sample fell somewhere within the range set by the error bars. He testified that while there were mechanisms and methodology to assist in determining where a given sample fell, such methodology was itself fraught with error and could not be applied reliably based upon the available information. TX 1340; Trial Tr. 1046-48 (Apley).

Moreover, Kodak had itself investigated the accuracy with which Mr. Van Dam could measure grains. In that analysis, Kodak concluded that the grains could by measured to an accuracy of 0.022 µ. Hence, for those products having a thickness of 0.2 µ, the measurement error was as much as about ± 10%. TX 1295 (EK129989-991), 1341 (¶ 52); Trial Tr. 151-53 (Van Dam). While the potential error determined by Kodak and the error ranges calculated by Dr. Apley differ, Kodak's study makes clear that consideration needs to be given to error analysis when assessing infringement.

Finally, Kodak objected to proposed TX 1386 and testimony concerning the effects of random sampling error (Source 1 error) and random measurement error (Source 2 error) on Agfa's data. Trial Tr. at 1024-27. Dr. Apley's report (TX 1340) opined only as to error factors relating to Kodak grain measurements. At deposition Dr. Apley informed Kodak that he had not performed his error analysis with respect to Agfa measurements. The night before Dr. Apley's testimony Agfa produced proposed Trial Exhibit 1386 setting forth an error analysis data of Agfa's data. Id. Kodak's objection to that exhibit was sustained, and Dr. Apley was not permitted testify about that analysis. Id. at 1026-27.

For these reasons, I recommend that Dr. Apley's bootstrap analysis of Kodak grain measurement data as set forth in TX 1340 be utilized and his margin of error boxes considered in the determination of infringement. In view of Dr. Apley's bootstrap analysis and the above discussed issues of non-homogeneity and small sample size relative to total product coated, it is my recommendation that it is appropriate for the Court to recognize the limits of accuracy in the use of TEMS for the acquisition of grain data. See In re Gabapentin Patent Litig., 393 F.Supp.2d 278 (D.N.J. 2005) ("[t]he pH levels are simply not precise enough for determining literal infringement"); Kimberly-Clark Corp. v. Tyco Healthcare Retail Group, 456 F.Supp.2d 998, 1005 (E.D.Wis. 2006) (the issue of how to treat measurement errors and variations arising in standard industry testing is a matter of fact left for the infringement stage); see also San Huan New Materials High Tech Inc. v. Int'l Trade Comm'n, 161 F.3d 1347, 1361 (Fed. Cir. 1998) (it "was not shown to be error, legal or scientific, . . . to recognize these limits of accuracy, and to round the measured weight percentages to the nearest integer")

Literal Infringement by the Accused Products CRONEX 10T

Kodak alleges that Cronex 10T infringes claims 1-3, 5, and 6 of the `426 patent. Independent claim 1 recites that the tabular silver halide grains having a thickness of less than 0.2 micron and an average aspect ratio of from 5:1 to 8:1 account for at least 50 percent of the total projected area of the silver halide grains present in the silver halide emulsion. Claims 2, 3, 5, and 6 depend from claim 1.

The grain size parameters for Cronex 10T were determined as set forth in Table I below for the identified tested film samples:

Table I (Cronex 10T) 426(1) Mfg. Batch Number AR % PA Date TX No. 578 6011 0070 01 5.67 5.46% Jan-96 TX-1305 (¶¶ EK107404, A541991), 1366, 1341 (¶¶ 102-7), 1368 578 6111 0184 03 6.40 6.13% Nov-96 TX-1305 (EK107412, A591767), 1366, 1341 (¶¶ 102-7), 1368 578 8051 0432 04 5.20 46.84% May-98 TX-1320 (EK105236, EK005268-70), 1368, 1341 (¶¶ 102-7), 1368 578 9031 0046 4.68 28.27% Mar-99 TX-1305 (¶¶ A534991), 1366, 1341 (¶¶ 102-7) 578 0081 0564 03 4.72 16.42% Aug-00 TX-1305 (¶¶ A538607, A541980), 1368, 1341 (¶¶ 102-7) 578 0091 0347 01 3.65 18.08% Sep-00 TX-1305 (EK111294, A582074), 1368, 1341 (¶¶ 102-7) 578 0111 0136 5.63 50.37% Nov-00 TX-1305 (¶¶ A527497, A534984), 1366, 1341 (¶¶ 102-7), 681 578 0111 0136 4.88 40.83% Nov-00 TX-1305 (¶¶ A527497, A541970), 1368, 1341 (¶¶ 102-7) 578 0129 441 4.91 17.92% Dec-00 TX-1305 (¶¶ A534985), 1368, 1341 (¶¶ 102-7) 578 1049 0539 4.87 31.76% Apr-01 TX-1305 (¶¶ A534986), 1368, 1341 (¶¶ 102-7) 578 1097 0278 01 4.46 54.23% Sep-01 TX-1305 (¶¶ A512775, A534965), 1368, 1341 (¶¶ 102-7) 578 1097 0278 01 4.47 54.12% Sep-01 TX-683, 1366, 1305 (A512775) 578 1107 0085 01 5.37 57.43% Oct-01 TX-684, 1366, 1305 (A512765) 578 1107 0085 01 5.38 50.38% Oct-01 TX-1320 (¶¶ A512765, EK107483-86), 1368, 1341 (¶¶ 102-7) 578 1117 0079 5.25 29.45% Nov-01 TX-1305 (¶¶ A534987), 1368, 1341 (¶¶ 102-7) 578 2028 0031 4.61 40.51% Feb-02 TX-1305 (¶¶ A527494, A534988), 1368, 1341 (¶¶ 102-7) 578 2034 0338 01 4.80 48.24% Mar-02 TX-1305 (¶¶ A512764, A534949), 1368, 1341 (¶¶ 102-7) 578 2057 0314 8.96 9.32% May-02 TX-1305 (¶¶ A527496, A534989), 1368, 1341 (¶¶ 102-7) 578 2067 0161 01 3.97 18.39% Jun-02 TX-1305 (¶¶ A534950, A512774), 1368, 1341 (¶¶ 102-7) 578 2087 0199 5.69 54.56% Jul-02 TX-325, 1320 (A527495), 1377 578 2087 0199 4.19 3.87% Jul-02 TX-1305 (¶¶ A527495, A534990), 1368, 1341 (¶¶ 102-7)

The above table includes four samples that would seem to literally infringe claim 1 of the `426 patent. Dr. Apley calculated an "error box" to 95% joint confidence for one of these samples, i.e., the sample possessing an average aspect ratio of 5.38 and a projected area of 50.38% taken from batch 5781107008501. TX 1340 (¶¶ 35-36); Trial Tr. at 1035-36 (Apley). Applying Dr. Apley's "error box" to the sample of 5781107008501 establishes that no conclusion can be reached with confidence as to whether the data point of the sample in fact falls within the claimed region. TX 1340 (¶ 35-36); Trial Tr. at 1035-36 (Apley). The data point is just as likely to fall outside the literal scope of claim 1 of the `426 patent as it is to fall within the literal scope.

Dr. Apley's report (TX 1340) did not contain a similar error analysis for the sample of batch 5781107008501 having an average aspect ratio of 5.37 and a projected area of 57.43%. Agfa also has not furnished error boxes for the samples of batches 57801110136 and 57820870199 which meet the literal size requirements of claim 1 of the `426 patent.

The results tabulated in Table I above are so scattered and irreconcilable with one another as to make their reliability dubious and their accuracy questionable.⁷ For example, the above-discussed batch 57820870199 produced one infringing sample having an aspect ratio of 5.69 and a projected area of 54.56%, and another non-infringing sample having an aspect ratio of 4.19 and a projected area of 13.87%. The above-discussed batch 57801110136 also produced a non-infringing sample that failed to meet both the average aspect ratio limitation and the projected area limitation. This leaves one sample of batch 578110700851 which was neither subject to Dr. Apley's analysis nor shared a non-infringing sample. Nonetheless, the occurrence of this sample likewise is in question, in view of Dr. Apley's testimony regarding sampling error and the Kodak conclusion that Mr. Van Dam's measurements were accurate to ± 0.022 µ. Further casting doubt on this sample is the vast majority of the samples and batches that do not fall within the literal scope of claim 1 of the `426 patent. Many of the samples and batches fall far outside the breadth of the claim. The aspect ratios of the samples range from 3.65 to 8.96, and several of the samples have projected areas of less than 10%. While some minor variations due to non-homogenity issues would have been expected, these variations are substantial. These measurements lack the reliability and consistency necessary to support a finding by a preponderance of literal infringement as to Cronex 10T in general. The variability of results and high percentage of non-infringing samples lessens to insignificance any confidence about accuracy of the would-be infringing sample. Astra Aktiebolag v. Andrx Pharm., 222 F.Supp.2d 423, 521-22 (S.D.N.Y. 2002), aff'd, In re Omeprazole Patent Litig., 84 Fed.Appx. 76 (Fed. Cir. 2003). Kodak was required to establish infringement by a preponderance, and the occurrence of an isolated sample out of so many cannot be deemed to be a preponderance of evidence regarding Cronex 10T in general or even that sample in particular.

7 While Kodak argues that the product is homogeneous, the variability in the aspect ratio — from 3.65 to 8.96- and percent projected area — from 5.46 to 57.43-suggest quite the contrary.

It is my recommendation that Kodak has not shown by a preponderance of the evidence that Cronex T infringes claim 1 of the `426 patent. Because claims 2, 3, 5, and 6 depend from claim 1, they also are not infringed. See, Wahpeton Canvas Co., Inc. v. Frontier, Inc., 870 F.2d 1546, 1553 n. 9 (Fed. Cir. 1989).

CRONEX 10TL

Kodak alleges that Cronex 10TL infringes claims 1-3, 5, and 6 of the `426 patent. Claims 2, 3, 5, and 6 depend from independent claim 1.

The grain size parameters for Cronex 10TL were determined as set forth in Table II below for the identified tested film samples: Grain Size Data Summary — Cronex 10TL Product Batch 426(1) Mfg. Line Number AR % PA Date TX No.

Table II (Cronex 10TL) Cronex 10TL 579 9071 0401 01 5.27 53.39% Jul-99 TX-1321 (EK107413, EK065550-54), 1368, 1341 (¶¶ 108- 12), 326 Cronex 10TL 579 9071 0401 01 4.94 66.30% Jul-99 TX-1306 (EK107413, A591765), 1368, 1341 (¶¶ 108-12) Cronex 10TL 579 0051 0309 04 4.49 47.62% May-00 TX-1306 (A534980, A541971), 1368, 1341 (¶¶ 108-12) Cronex 10TL 579 0091 0250 01 4.40 57.93% Sep-00 TX-1306 (EK107409, A591766), 1368, 1341 (¶¶ 108-12), 688 Cronex 10TL 579 1041 0117 01 5.15 71.37% Apr-01 TX-689, 1366, 1306 (A512794) Cronex 10TL 579 1088 0432 01 5.32 62.83% Aug-01 TX-1306 (A512787, A591788), 1366, 1341 (¶¶ 108-12), 691 Cronex 10TL 579 1088 0432 01 5.24 64.96% Aug-01 TX-1321 (A512787, EK107529-31), 1368, 1341 (¶¶ 108- 12), 327 Cronex 10TL 579 2018 0329 02 4.85 53.80% Jan-02 TX-1306 (A512771, A541478), 1368, 1341 (¶¶ 108-12), 692 Cronex 10TL 579 2047 0212 01 5.18 79.35% Apr-02 TX-1306 (A512767, A541477), 1366, 1341 (¶¶ 108-12), 695 Cronex 10TL 579 2047 0212 01 5.19 79.42% Apr-02 TX-694, 1366, 1306, (A512767) Cronex 10TL 579 2067 0292 01 5.20 83.31% Jun-02 TX-1306 (A512770, A541476), 1366, 1341 (¶¶ 108-12), 697 Cronex 10TL 579 2067 0292 01 5.22 83.33% Jun-02 TX-696, 1366, 1306, (A512770) Cronex 10TL 579 2087 0150 01 4.78 61.48% Aug-02 TX-1306 (A512769, A541475), 1368, 1341 (¶¶ 108-12) Cronex 10TL 579 2087 0150 01 4.79 61.63% Aug-02 TX-698, 1366, 1306, (A512769) Dr. Apley's Report, TX 1340, includes calculations of grain size properties of film samples of batches 5799071040101 and 5791088043201. TX 1340 (¶¶ 31-34). The Report also includes these properties plotted on a graph with their 95% joint confidence region as calculated by Dr. Apley. TX 1340 (¶¶ 31-34); Trial Tr. at 1036 (Apley). Applying Dr. Apley's "error boxes" to the allegedly "literal infringement" grain data points establishes that no conclusion can be reached with confidence as to whether the data points in fact fall within the claimed region. TX 1340 (¶¶ 31-34); Trial Tr. at 1036 (Apley). Neither of these film samples can be said to more likely fall within the literal scope of claim 1 of the `426 patent than outside of the literal scope when the error is taken into account. TX 1340 (¶¶ 31-34); Trial Tr. at 1036 (Apley).

Dr. Apley's report did not contain a similar error analysis for the sample of batch 5791088043201 having an average aspect ratio of 5.32 and a projected area of 62.83%. Agfa also has not provided error boxes for any of the following: the sample of batch 5791041011701 having an average aspect ratio of 5.15 and a projected area of 71.37%; the samples of batch 5792047021201 having average aspect ratios of 5.18 and 5.19 and projected areas of 79.35% and 79.42%, respectively; and the samples of batch 5792067029201 having average aspect ratios of 5.20 and 5.22 and projected areas of 83.31% and 83.33%, respectively. Each batch meets the literal size requirements of claim 1 of the `426 patent.

Unlike the scattered and discrepant data points relating to the Cronex 10T product, the data points of Table II are sufficiently consistent as to reasonably lead to a conclusion of their reliability. Further, the percent projected areas in Table II are much higher than the values in Table I, suggesting a greater likelihood that error did not lead to a false conclusion of infringement. Accordingly, it is my recommendation that Cronex 10TL literally infringes claim 1 of the `426 patent. Grain Processing Corp. v. American Maize-Prods., 893 F.Supp. 1386, 1395-96 (N.D.Ind. 1995), damages award vacated and remanded, affirmed in other respects, 1997 WL 71726, 1997 U.S. LEXIS 2885 (Fed. Cir. 1997). It is my further recommendation that Cronex 10TL literally infringes claims 2, 3, 5 and 6. See Depo. Desig. Tab 17 (07/08/04 M. Guy) at 206-07 (the support is a blue tinted, transparent film support) (claims 2 and 3); Depo. Desig. Tab 17 (07/08/04 M. Guy) at 206 (the dispersing medium is a hardenable hydrophilic colloid) (claim 5); Depo. Desig. Tab 18 (10/07/04 M. Guy) at 241 (the silver halide is silver bromide) (claim 6).

The extent of infringement, i.e., the percentage of the Cronex 10TL production which infringes the `426 patent, has yet to be determined and normally would be decided during the damages phase. Nickson Indus., Inc. v. Rol Mfg. Co., Ltd., 847 F.2d 795, 798-99 (Fed. Cir. 1988) (Obligation of infringer to establish quantum of non-infringing sales when assessing damages). However, in view of my recommendation above regarding Kodak's failure to provide notice under Section 287 prior to the expiration of the `426 patent, Kodak is not entitled to damages for infringement by Cronex 10TL.

ULTRAVISION C

Kodak alleges that Ultravision C infringes claims 1-3, 5, and 6 of the `426 patent. Claims 2, 3, 5, and 6 depend from independent claim 1.

The grain size parameters for Ultravision C were determined as set forth in Table III below for the identified tested film samples. Grain Size Data Summary — Ultravison C Product Batch 426(1) Mfg. Line Number AR % PA Date TX No. Grain Size Data Summary-Curix Ultra UV-C Never Sold in U.S. Prior to Patent Expirations

Table III Ultravision C 565 7103 0653 01 4.40 26.88% Oct-97 TX-1307 (EK105239, A541962), 1368, 1341 (¶¶ 135- 6) Ultravision C 565 0081 0376 01 5.13 27.70% Aug-00 TX-1307 (EK111270, A582083), 1368, 1341 (¶¶ 135- 6) Ultravision C 565 1051 0592 01 4.93 32.25% May-01 TX-1307 (A512800, A534945), 1368, 1341 (¶¶ 135-6) Ultravision C 565 1088 0394 01 4.93 45.22% Aug-01 TX-1307 (A512790, A534946), 1368, 1341 (¶¶ 135-6) Ultravision C 565 1088 0394 01 5.84 30.66% Aug-01 TX-1322 (A512790, EK107575-76), 1368, 1341 ¶¶ 135-6) Ultravision C 565 1097 0434 01 4.81 12.60% Sep-01 TX-1307 (A527487, A591786), 1368, 1341 (¶¶ 135-6) Ultravision C 565 2017 0204 01 4.82 26.64% Jan-02 TX-1307 (A512805, A534962), 1368, 1341 (¶¶ 135-6) Ultravision C 565 2037 0529 01 4.80 33.29% Mar-02 TX-1307 (A512804, A534963), 1368, 1341 (¶¶ 135-6) Ultravision C 79880066 5.82 4.38% Apr-02 TX-1307 (A503446, A617519), 1368, 1341 (¶¶ 135-6) Ultravision C 565 2067 0256 01 5.09 22.39% Jun-02 TX-1307 (A512798, A534947). 1368, 1341 (¶¶ 135-6) Ultravision C 565 2087 0106 01 4.80 29.87% Aug-02 TX-1307 (A512806, A534964), 1368, 1341 (¶¶ 135-6) Curix Ultra UV-C 79880066 5.82 4.48% Apr-02 TX-799, 1307, (A503446, A617519) None of these film samples can fall within the literal scope of claim 1 of the `426 patent because none literally satisfy the "at least 50 percent of the total projected area" limitation of the claim, and most do not satisfy the average aspect ratio limitation of between 5:1 and 8:1. My recommendation is that Ultravision C does not literally infringe claims 1-3 and 5-6 of the `426 patent.

ULTRAVISION Ci

Kodak alleges that Ultravision Ci infringes independent claim 1 and dependent claims 2, 3, and 5-7 of the `425 patent, and independent claim 1 and dependent claims 2, 4, 5, 8, 9, 13, 16, and 28 of the `520 patent.

The grain size parameters for Ultravision Ci were determined as set forth in Table IV below for the identified tested film samples. Grain Size Data Summary — Ultravision Ci Product Batch 425/520(1) 425/520(19) Mfg. Line Number AR % PA AR %PA Date TX No.

Table IV Ultravision Ci 564 8111 4.34 69.89% 4.34 69.89% Nov- TX-1308 (EK105240, A541961), 1368, 1341 (¶¶ 95- 0358 01 98 100) Ultravision Ci 560 0021 5.72 88.36% 5.66 89.24% Feb- TX-1308 (EK111268, A582094), 1368, 1341 (¶¶ 95- 0266 04 00 100) Ultravision Ci 560 0021 4.96 74.31% 4.91 85.76% Feb- TX-1308 (EK111282, A582093), 1368, 1341 (¶¶ 95- 0267 01 00 100) Ultravision Ci 560 0021 4.81 85.98% 4.81 85.98% Feb- TX-1308 (EK111283, A582092), 1368, 1341 (¶¶ 95- 0268 01 00 100)

Kodak also points to Trial Exhibit 345A-D as providing grain measurements for an emulsion in which grains with a thickness of less than 0.3 micron and a diameter of greater than or equal to 0.6 micron provide 92.10% of the total projected area and have an average aspect ratio of 7.09:1 (TX 345A-D) None of these film samples can be said to fall within the literal scope of claim 1 or claim 19 of the `425 and `520 patents because none literally satisfy the average aspect ratio greater than 8:1 requirement of the claims.

Kodak also alleges that Ultravision Ci infringes independent claim 1 and dependent claims 2, 3, 5, and 6 of the `426 patent. Trial Exhibit 800 sets forth Agfa generated data for an emulsion in which grains with a thickness of less than 0.2 micron provide 57.06% of the total projected area and have an average aspect ratio of 4.73:1 (TX 800) This emulsion does not fall within the literal scope of claim 1 of the `426 patent because it does not literally satisfy the average aspect ratio of from 5:1 to 8:1.

It is therefore my recommendation that Ultravision Ci does not infringe literally any of the asserted claims of the `426, `425, and `520 patents.

ULTRAVISION L

Kodak alleges that Ultravision L infringes independent claim 1 and dependent claims 2, 3, 5, and 6 of the `426 patent, independent claim 1 and dependent claims 2, 3, and 5-7 of the `425 patent, and independent claim 1 and dependent claims 2, 4, 5, 8, 9, 13, 16, and 28 of the `520 patent.

The grain size parameters for Ultravision L were determined as set forth in Table V below for the identified tested film samples. Grain Size Data Summary — Ultravision L Product Batch 426(1) 425/520 (1) 425/520 (19) Mfg. Line Number AR %PA AR %PA AR %PA Date TX No. Grain Size Data Summary — Curix Ultra UV-L Never Sold in U.S. Prior to Patent Expirations

Table V TX-1309 (EK105241, A541963),1368,1341 (¶¶ 72- Ultravision L 442917 4.63 64.92% 4.27 93.53% Mar-98 78, 133-4) 564 8091 0442 TX-1309 (A582084, EK111265), 1368, 1341 (¶¶ 72- Ultravision L 01 4.49 38.94% 4.16 95.65% Sep-98 78, 133-4) 564 9121 0169 TX-1309 (A534976, A541969), 1368, 1341 (¶¶ 72- Ultravision L 01 4.81 51.90% 4.56 96.62% Dec-99 78, 133-4) 564 0031 0838 TX-1309 (A534978, A541968), 1368, 1341 (¶¶ 72- Ultravision L 04 4.54 51.60% 4.55 89.97% Mar-00 78, 133-4) 564 0061 0275 TX-1309 (EK111267, A582086), 1368, 1341 (¶¶ 72- Ultravision L 01 5.00 4.84% 4.69 52.15% 4.13 95.04% Jun-00 78, 133-4) 564 0091 0101 TX-1309 (EK111266, A582085), 1368, 1341 (¶¶ 72- Ultravision L 01 4.67 57.14% 4.64 97.43% Sep-00 78, 133-4) 564 1051 0567 TX-1309 (A512796, A535121). 1368, 1341 (¶¶ 72- Ultravision L 01 1.26 0.03% 5.37 83.48% 5.29 97.72% May-01 78, 133-4) 564 1078 0364 TX-1309 (A512792, A535122), 1368, 1341 (¶¶ 72- Ultravision L 01 4.87 61.33% 4.61 92.64% Jul-01 78, 133-4) 564 1088 0368 TX-1309 (EK107411; EK111292, A591770), 1368, Ultravision L 01 4.49 32.86% 4.08 92.61% Aug-01 1341 (¶¶ 72-78, 133-4) 564 1097 0136 TX-1309 (A527490, A591781), 1368, 1341 (¶¶ 72- Ultravision L 01 4.42 69.83% 4.24 93.54% Sep-01 78, 133-4) 564 1127 0140 TX-1309 (A512791, A535123), 1368, 1341 (¶¶ 72- Ultravision L 01 4.90 69.44% 4.72 95.61% Dec-01 78, 133-4) 564 2037 0449 TX-1309 (A512793, A535124), 1368, 1341 (¶¶ 72- Ultravision L 02 1.39 0.04% 4.83 58.25% 4.99 96.81% Mar-02 78, 133-4) 564 2067 0215 TX-1309 (A512772, A534941), 1368, 1341 (¶¶ 72- Ultravision L 01 4.70 56.44% 4.31 90.68% Jun-02 78, 133-4) 564 2087 0067 TX-1309 (A512789, A534935), 1368, 1341 (¶¶ 72- Ultravision L 01 7.51 3.93% 4.75 73.74% 4.50 92.15% Aug-02 78, 133-4) 564 2097 0255 TX-1309 (A512788, A534943), 1368, 1341 (¶¶ 72- Ultravision L 01 4.43 62.15% 4.08 95.17% Sep-02 78, 133-4) Grain Size Data Summary — Ultravision L Product Batch 426(1) 425/520(1) 425/520 (19) Mfg. Line Number AR %PA AR %PA AR %PA Date TX No. Product Batch 426(1) 425/520(1) 425/520(19) Mfg. Line Number AR %PA AR %PA AR %PA Date TX No. Curix Ultra UV-L 40775065 9.88 2.10% 8.60 98.13% 8.43 99.29% May-01 TX-801, 1310, (A503448, A534939) Curix Ultra UV-L 40775065 8.60 98.20% 8.43 99.29% May-01 TX-803, 1366, 1310, (A503448) With the exception of product sample 40775065 (discussed below), none of these film samples can be said to fall within the literal scope of claim 1 of the `25, `426, and '520 patents or claim 19 of the `425 and `520 patents. The film samples all fail to satisfy one or both of the "at least 50 percent of the total projected area" and the average aspect ratio of 5:1 to 8:1 limitations of the `426 patent claims. Additionally, each of the film samples fails to literally satisfy one or both the "at least 50 percent of the total projected area" and the "average aspect ratio of greater than 8:1" limitations of the claims of the `425 and `520 patents. TX 1041 (¶¶ 72-78); Trial Tr. at 1040 (Apley).

The unrebutted testimony is that Ultravision L product sample 40775065 was an experimental product made in Belgium using a different process than the Ultravision L products that were used, manufactured, offered for sale or sold in the United States. TX 1341 (¶¶ 72). This testimony is supported by the remainder of Kodak's and Agfa's grain measurements of the Ultravision L product. These grain measurements demonstrate that the Ultravision L product made or sold in the United States consistently falls outside of the literal scope of the T-grain patents. TX 1341 (¶ 72).

Even if the Ultravision L product sample 40775065 was made, used, or sold in the United States, this isolated data point would not by a preponderance of the evidence be sufficient to find the Ultravision L product line infringes literally any of the T-grain patents. The isolated data point would be considered an anomaly when considered in reference to the other data. It therefore is my recommendation that Ultravision L does not infringe literally the T-grain patents.

ORTHOVISION G

Kodak alleges that Orthovision G infringes independent claim 1 and dependent claims 2, 3, 5, and 6 of the `426 patent, independent claim 1 and dependent claims 2, 3, and 5-7 of the `425 patent, and independent claim 1 and dependent claims 2, 4, 5, 8, 9, 13, 16, and 28 of the `520 patent.⁸

8 Considering the arguments made to the PTO in distinguishing the prior art by virtue of grain size the aspect ratio, among other attributes, it is surprising that Kodak would accuse a single product of infringing all of the T-grain patents.

The grain size parameters for Orthovision G were determined as set forth in Table VI below for the identified tested film samples. Grain Size Data Summary — Orthovision G Product Batch 426(1) 425/520(1) Mfg. Line Number AR % PA AR % PA Date TX No.

Table VI Orthovision G 525 9043 6.17 40.29% 5.70 94.50% Apr-99 TX-1311 (EK015237, A541957), 1368, 1341 0152 02 (¶¶ 88-91, 113-20) Orthovision G 525 9043 7.59 75.62% 7.15 97.30% Apr-99 TX-1325 (EK015237, EK005259-61), 1341 0152 02 (¶¶ 88-91, 113-20), 341 Orthovision G 525 9113 6.35 75.48% 6.10 96.25% Nov-99 TX-1311 (A538605, A541978), 1341 (¶¶ 88-91, 0134 02 113-20), 795 Orthovision G 525 9113 6.48 76.38% 6.20 96.38% Nov-99 TX-1325 (A538605, EK112389-93), 1368, 0134 02 1341 (¶¶ 88-91, 113-20) Orthovision G 525 9113 6.11 41.65% 5.57 94.72% Nov-99 TX-1311 (A538606, A541979), 1368, 1341 0480 21 (¶¶ 88-91, 113-20) Orthovision G 5259093218 6.25 81.72% TX 793 With respect to claim 1 of the `426 patent, Dr. Apley's report plots the grain size properties of the film sample of Batch No. 5259043015202 having an average aspect ratio of 7.59 and a projected area of 75.62%. In addition, a 95% joint confidence region as calculated by Dr. Apley was plotted. TX 1340 (¶¶ 37-38). Applying Dr. Apley's "error box" to the allegedly "literal infringement" grain data point establishes that no conclusion can be reached with confidence as to whether the data point in fact falls within the claimed region. TX 1340 (¶¶ 37-38). The film sample cannot be said to more likely fall within the literal scope of claim 1 of the `426 patent when the error scope is taken into account. TX 1340 (¶¶ 37-38).

However, samples for batch 5259113013402 providing average aspect ratios of 6.35 and 6.48 and projected areas of 75.48% and 76.38%, respectively, fall squarely in the middle of the claimed ranges of the `426 patent. A sample for batch 5259093218 having an average aspect ratio of 6.25 and a projected area of 81.72% likewise falls in the claimed ranges of the `426 patent. The remaining two grain data points fall outside the claimed range.

For these reasons, it is my recommendation that Orthovision G be found to in-fringe literally at least claim 1 of the `426 patent. It is my further recommendation that Orthovision G be found to infringe literally claims 2, 3, 5, and 6 of the `426 patent. With respect to claims 2 and 3, the support of Orthovision G is a blue tinted transparent film. Depo. Desig. Tab 18 (10/07/04 M. Guy) at 246-47. With respect to claims 5 and 6, the dispersing medium is comprised of a hardenable hydrophilic colloid, and Orthovision G contains silver bromide. Id. at 241, 246-47. The determination of the quantum of infringement will need to be made during the damages phase, however.

The Orthovision G product does not literally infringe claim 1 of the `425 and `520 patents because none of the data points literally satisfies the average aspect ratio requirement of greater than 8:1 of claim 1 of the `425 and `520 patents.

ORTHOVISION L

Kodak alleges that Orthovision L infringes claims 1-3, 5, and 6 of the `426 patent. Claims 2, 3, 5, and 6 depend from independent claim 1.

The grain size parameters for Orthovision L were determined as follows in Table VII for the identified tested film samples: Grain Size Data Summary-Orthovision L Product Batch 426(1) Mfg. Line Number AR % PA Date TX No.

Table VII Orthovision L 527 8093 0510 02 6.09 21.70% Sep-98 TX-1312, (EK107399, A591773) 1368, 1341 (¶¶ 140- 1) Orthovision L 527 9033 0250 04 4.18 8.49% Mar-99 TX-1312 (EK105238, A541958), 1368, 1341 (¶¶ 140- 1) Orthovision L 527 0023 0194 03 4.97 50.11% Feb-00 TX-1312 (EK111295, A582072), 1368, 1341 (¶¶ 140- 1) None of these film samples fall within the literal scope of claim 1 of the `426 patent because none literally satisfy both the "at least 50 percent of the total projected area" limitation of the claim, and the average aspect ratio limitation of between 5:1 and 8:1. It is my recommendation that the Orthovision L product does not literally infringe the claims of the `426 patent.

AGFA CP-BU

Kodak alleges that Agfa CP-BU infringes independent claim 1 and dependent claims 2, 3, 5, and 6 of the `426 patent, independent claims 1 and 19 and dependent claims 2, 3, and 5-7 of the `425 patent, and independent claims 1 and 19 and dependent claims 2-5, 8-10, 13, 16, 20 and 28 of the `520 patent.

The grain size parameters for Agfa CP-BU were determined as follows in Table VIII for the identified tested film samples: Grain Size Data Summary — CX1X5 Product Batch 426(1) 425/520(1) 425/520(19) Mfg. Line Number AR %PA AR %PA AR %PA Date TX No.

Table VIII TX-1327 (EK10523, EK005311-12), 1368, 1341 CX1X5 466 9080 8611 01 8.51 45.83% 8.22 93.55% 7.14 97.69% Aug-99 (¶¶ 79-87, 130-2), 328 TX-1313 (EK105231, A541498), 1368, 1341 (¶¶ 79- CXIX5 466 9080 8611 01 8.56 4.01% 7.89 94.49% 7.57 98.52% Aug-99 87, 130-2), 700 TX-1313 (EK105232, A541489), 1368, 1341 (¶¶ 79- CP-BU 40605071 9.23 8.32% 7.15 95.65% 6.90 97.44% Dec-99 87, 130-2), 701 TX-1327 (EK105232, EK005295-97), 1368, 1341 CXIX5 40605071 8.26 22.83% 7.49 95.32% 7.05 99.52% Dec-99 (¶¶ 79-87, 130-2), 329 TX-1313 (EK105235, A541501), 1368, 1341 (¶¶ 79- Cronex 5 40605071 8.26 6.78% 7.13 94.29% 7.03 97.86% Dec-99 87, 130-2), 702 TX-1327(EK105235, EK005303-05), 1368, 1341 CXIX5 40605071 8.61 31.33% 8.06 93.93% 7.58 97.70% Dec-99 (¶¶ 79-87, 130-2), 330 TX-1313 (A503457, A534942), 1368, 1341 (¶¶ 79- CXIX5 39665071 8.46 6.35% 7.15 90.06% 6.81 96.02% Jun-00 87, 130-2), 704, 710 TX-1313 (A503458, A534970), 1368, 1341 (¶¶ 79- CXIX5 79775023 7.97 7.44% 7.41 97.26% 7.32 98.22% May-01 87, 130-2), 708, 709 TX-1313 (A512757, A535107), 1368, 1341 (¶¶ 79- Radiomat B Plus 466 0030 4504 21 7.43 22.71% 6.89 95.38% 6.81 98.21% Mar-00 87, 130-2) TX-1327(A512757, EK107454-56), 1368, 1341 CXIX5 466 0030 4504 21 9.36 17.98% 8.02 91.19% 7.77 96.91% Mar-00 (¶¶ 79-87, 130-2), 331 TX-1313 (A512760, A5351 19), 1368, 1341 (¶¶ 79- Radiomat B Plus 466 1060 3205 01 8.71 28.64% 7.77 94.04% 7.34 97.04% Jun-01 87, 130-2), 712 TX-1313 (A512759, A535108), 1368, 1341 (¶¶ 79- CXIX5 466 2010 4001 01 9.05 14.16% 7.88 97.78% 7.76 98.60% Jan-02 87, 130-2) TX-1313 (A512761, A541479), 1368, 1341 (¶¶ 79- Radiomat B Plus OB7029 8.78 10.00% 7.42 95.64% 7.42 95.64% Feb-02 87, 130-2) Grain Size Data Summary — CP-BU New Product Batch 426(1) 425/520(1) 425/520(19) Mfg. Line Number AR %PA AR %PA AR %PA Date TX No. TX-1313 (EK111293, A582057), 1368, 1341 (¶¶ 79- Agfa CP-BU New 79765033 8.33 0.89% 7.30 95.93% 7.10 97.63% Apr-00 87, 130-2) Agfa CP-BU New 79765033 8.33 0.89% 7.30 95.93% 7.10 97.63% Apr-00 TX-707, 1366, 1313, (EK111293) TX-1313 (EK111281, A582058), 1368, 1341 (¶¶ 79- Agfa CP-BU New 40675122 9.53 12.42% 7.35 96.27% 7.14 97.81% Jul-00 87, 130-2), 705 TX-1313 (EK111280, A582056), 1368, 1341 (¶¶ 79- Agfa CP-BU New 39705025 10.65 11.94% 7.50 97.70% 7.30 99.17% Oct-00 87, 130-2), 706 TX-1313 (EK111296, A582055), 1368, 1341 (¶¶ 79- Agfa CP-BU New 21787014 7.34 96.15% 7.24 98.38% Jun-01 87, 130-2), 711 None of these film samples fall within the literal scope of claim 1 of the `426 patent because none literally satisfy the "at least 50 percent of the total projected area" requirement of the claim, and many of the samples do not have an average aspect ratio of 5:1 to 8:1, as recited in claim 1 of the `426 patent. The products in Table VIII do not fall within the literal scope of the claims of the `426 patent.

The film samples also do not infringe literally claim 19 of the `425 patent or claim 19 of the `520 patent. Each claim 19 requires that at least 50 percent of the total projected area having a thickness of less than 0.5 micron, a diameter of at least 0.6 micron, and an average aspect ratio of greater than 8:1. None of the film samples is characterized by an average aspect ratio of greater than 8:1.

The vast majority of samples and batches do not fall within the literal scope of claim 1 of the `425 or `520 patents. There are three data points having silver halide grains with a thickness of less than 0.3 micron, a diameter of at least 0.6 micron, and an average aspect ratio of greater than 8:1 which seemingly meet the at least 50% total projected area limitation of claim 1 of the `425 patent and claim 1 of the `520 patent. Two of these data points, i.e., (a) a sample for batch 4669080861101 having an average aspect ratio of 8.22 and a projected area of 93.55% and (b) a sample for batch 40605071 having an average aspect ratio of 8.06 and a projected area of 93.93%, were subject to error analyses by Dr. Apley. The error analyses demonstrated that no conclusion can be reached with confidence as to whether the data points in fact fall within the claimed region. TX 1340 (¶ 27-30; 47-48).

Although no error analysis was conducted for the remaining data point sample of batch number 4660030450421, the sample had an aspect ratio of 8.02 and a projected area of 91.19%. The aspect ratio of the sample meets the "at least 8:1" limitation of the `425 and `520 patents by only two hundredths. The reliability and accuracy of this data point is brought into serious doubt, particularly by Dr. Apley's other data analyses and Kodak's determination that Mr. Van Dam was able to measure to within ± 0.022 u. Further, this batch also produced a non-infringing sample. Specifically, batch 4660030450421 produced a sample having an aspect ratio of 6.89 and a projected area of 95.38. The same observation can be made for the other two data points which might be deemed infringing. Batch 4669080861101 produced a non-infringing sample having an aspect ratio of 7.89 and a projected area of 94.49. Batch 40605071 produced multiple non-infringing samples having average aspect ratios of less than 8:1. The fact that samples from the same batch can have such differing physical parameters supports my belief that error analysis is required.

The few data points that fall within the grain size limitations lack the reliability and consistency necessary to support a finding of literal infringement. Further, Dr. Apley has demonstrated that as to two of these data points no conclusive finding can be made about literal infringement. Furthermore, the variability of results and the small sample relative to total product coated lessens to insignificance any confidence that the batch literally infringes. It is therefore my recommendation that no literal infringement be found with respect to the Agfa CP-BU product against any of the asserted claims.

CURIX ORTHO HT-G

Kodak alleges that Curix Ortho HT-G infringes claims 1-3 and 5-8 of the `426 patent. Claims 2, 3, and 5-8 depend from independent claim 1.

The grain size parameters for Curix Ortho HT-G were determined as follows in

Table IX for the identified tested film samples: Grain Size Data Summary-HT-G Product Batch 426 (1) Mfg. Line Number AR % PA Date TX No.

Table IX Ortho HT-G 40410030 5.61 39.38% Jun-97 TX-1314 (A131835-36), 1368, 1341 (¶¶ 137-9) Ortho HT-G 39445055 6.12 7.17% May-98 TX-1314 (A131819-20), 1368, 1341 (¶¶ 137-9) Ortho HT-G 59435062 7.63 3.23% Jul-99 TX-1314 (EK105234, A541495), 1368, 1341 (¶¶ 137- 9) Ortho HT-G 59435062 7.37 51.62% Jul-99 TX-1328 (EK105234, EK005285-87), 1368, 1341 (¶¶ 137-9), 333 Ortho HT-G 40615051 6.98 10.30% Jan-00 TX-733, 1314, (A503435, A534937), 1368 Ortho HT-G 59635021 5.32 10.17% Mar-00 TX-1314, (A527478, A541981), 1368, 1341 (¶¶ 137-9) Ortho HT-G 40685042 5.73 5.58% Apr-00 TX-1314 (EK111279, A582078), 1368, 1341 (¶¶ 137- 9) Ortho HT-G 40665028 6.35 17.97% Jun-00 TX-1314 (A503436, A534972; A131802-03), 1368, 1341 (¶¶ 137-9) Ortho HT-G 50675065 8.75 25.06% Jul-00 TX-1314 (A534961, A512799), 1368, 1341 (¶¶ 137-9) Ortho HT-G 50675066 6.50 20.11% Jul-00 TX-1314 (EK11127, A582054), 1368, 1341 (¶¶ 137-9) Ortho HT-G 39305101 6.78 7.12% Aug-00 TX-1314 (A131768-69), 1368, 1341 (¶¶ 137-9) Ortho HT-G 59745006 6.64 19.34% Feb-01 TX-1314 (A512786, A534967), 1368, 1341 (¶¶ 137-9) Ortho HT-G 59745006 7.11 24.54% Feb-01 TX-1328 (A512786, EK107514-16), 1368, 1341 (¶¶ 137-9) Ortho HT-G 425 1040 4812 01 4.19 0.49% Apr-01 TX-1314 (A541983, A527480), 1368, 1341 (¶¶ 137-9) Ortho HT-G 425 1070 1208 01 5.18 3.57% Jul-01 TX-1314 (A582098, EK111288), 1368, 1341 (¶¶ 137- 9) Ortho HT-G 425 1070 8413 01 5.09 0.79% Jul-01 TX-1314 (A582097, EK111287), 1368, 1341 (¶¶ 137- 9) Ortho HT-G 425 1070 8602 01 6.92 7.93% Jul-01 TX-1314 (A534940, A512776), 1368, 1341 (¶¶ 137-9) Ortho HT-G 59850018 5.62 2.89% Jan-02 TX-1314 (A512777, A534951), 1368, 1341 (¶¶ 137-9) Ortho HT-G 79880001 7.73 8.94% Apr-02 TX-1314 (A503439, A534974, A131785-86), 1368, 1341 (¶¶ 137-9) RadiomatSG OH2011 8.32 26.28% Aug-02 TX-1314 (A512750, A535128), 1368, 1341 (¶¶ 137-9) A single film sample falls within the literal scope of claim 1 of the `426 patent. All of the other data points do not literally satisfy the "at least 50 percent of the total projected area'" requirement of the claim. The single data point is so isolated and remote from the remainder of the data that it cannot be considered to represent the product or to be free from error. It is my recommendation that the Court find that Ortho HT-G does not literally infringe the `426 patent claims.

DI IMAGE PLUS GREEN

Kodak alleges that DI Image Plus Green infringes claims 1-3, 5 and 6 of the `426 patent. Claims 2, 3, 5 and 6 depend from independent claim 1.

The grain size parameters for DI Image Plus Green were determined as follows in

Table X below for the identified tested film samples: Grain Size Data Summary — Curix Ortho DGGV2 Product Batch 426(1) Mfg. Line Number AR % PA Date TX No.

Table X Curix Ortho 39305101 6.78 7.12 Jun-97 TX-713, 1366, 1368 DGGV2 DI Image Plus QB3131 5.16 17.95% Feb-00 TX-1315 (A538608, A541973), 1368, 1341 (¶¶ 142-4) Green DI Image Plus 439 0060 1311 21 5.38 22.94% Jun-00 TX-1315 (A512801, A541481), 1368, 1341 (¶¶ 142-4) Green Curix Ortho 39665013 6.53 27.24% Jun-00 TX-1314 (A503441, A534971), 1366, 1341 (¶¶ 142- DGGV2 4), 1368 Curix Ortho 39665013 5.68 19.36% Jun-00 TX-1314 (A503441, A534959, A131752-53), 1366, DGGV2 1341 (¶¶ 142-4), 1368 Curix Ortho 39665013 6.82 27.70% Jun-00 TX-1314 (A503441, A534960), 1366, 1341 (¶¶ 142-4), DGGV2 1368 DI Image Plus 439 0070 0701 05 5.27 20.39% Jul-00 TX-1315 (A527492, A591785), 1368, 1341 (¶¶ 142-4) Green DI Image Plus 439 0110 5409 01 5.58 25.96% Nov-00 TX-1315 (A527491, A591784), 1368, 1341 (¶¶ 142-4) Green DI Image Plus 439 0120 3801 21 4.76 8.50% Dec-00 TX-1315 (A512753, A541487), 1368, 1341 (¶¶ 142-4) Green DI Image Plus 439 0120 3801 21 5.87 35.13% Dec-00 TX-1329 (A512753, EK107437-40), 1368, 1341 Green (¶¶ 142-4) Curix Ortho 79755038/08 5.04 19.55% Mar-01 TX-1314 (A503442, A131737-38), 1366, 1341 DGGV2 (¶¶ 142-4), 1368 Curix Ortho 79790022/02 5.88 24.49% Jul-01 TX-1314 (A503443, A534973, A131724-25), 1366, DGGV2 1341 (¶¶ 142-4), 1368 Curix Ortho 79790022/02 5.73 26.96% Jul-01 TX-1314 (A503443, A534969, A541480), 1366, 1341 DGGV2 (¶¶ 142-4), 1368 Radiomat G Plus 439 1080 4315 04 5.64 56.62% Aug-01 TX-1315 (A512756, A541500), 1366, 1341 (¶¶ 142-4), 1368, 726 Radiomat G Plus 439 1080 4315 04 4.70 33.12% Aug-01 TX-1315 (A512756, A541499), 1368, 1341 (¶¶ 142-4) Radiomat G Plus 439 2010 5810 02 5.67 6.40% Jan-02 TX-1315 (A541486, A512755), 1368, 1341 (¶¶ 142-4) Radiomat G Plus OC3020 5.02 23.89% Mar-02 TX-1315 (A512747, A541488), 1368, 1341 (¶¶ 142-4) DI Image Plus OG2059 7.30 6.75% Jul-02 TX-1315 (A512749, A582099), 1368, 1341 (¶¶ 142-4) Green One of the samples reported in Table X falls seemingly within the grain size limitations of claim 1 of the `426 patent. Batch 4391080431504 produced a sample having an average aspect ratio of 5.64 and a projected area of 56.62%. Dr. Apley's report TX 1340 did not contain an error analysis for the sample. However, the same batch 4391080431504 produced a sample having an average aspect ratio of 4.70 and a projected area of 33.12%, each falling outside the claimed ranges of the `426 patent. Additionally, the remaining samples reported in Table X do not literally infringe claim 1 of the `426 patent, and many of the samples and batches fall significantly far outside the breadth of the claim. For example, projected areas of the non-infringing data points range between 6.40% and 35.13%, well outside the required 50% level of claim 1. The one infringing data point appears to be an anomaly or in error, lacking the reliability and consistency necessary to support a finding of literal infringement. I therefore recommend that Curix Ortho DGGV2 be found to not literally infringe.

OPTHOS H

Kodak alleges that Opthos H infringes independent claim 1 and dependent claims 2, 3, and 5-8 of the `426 patent, independent claim 1 and dependent claims 2, 3, and 5-9 of the `425 patent, and independent claim 1 and dependent claims 2-5, 8, 10, 13, 16, and 28 of the `520 patent.

The grain size parameters for Opthos H were determined as follows in Table XI below for the identified tested film samples: Grain Size Data Summary — Opthos H Product Batch 426(1) 425/520(1) Mfg. Line Number AR % PA AR % PA Date TX No.

Table XI Curix Opthos 39405039 3.37 2.68% 4.22 79.24% Apr-98 TX-1316 (EK111274, A582066), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 39405039 4.42 93.60% Apr-98 TX-1316 (EK111274, A582061), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 59315075 7.19 97.04% Jul-98 TX-338 (EK065465-67) H LS/HC Curix Opthos 59315075 7.49 70.29% Jul-98 TX-337 (EK065488-90) H HS/LC Curix Opthos 39625050 4.36 1.61% 5.20 90.98% Feb-00 TX-1316 (EK111273, A582059), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 39625050 4.82 93.92% Feb-00 TX-1316 (EK111273, A582062), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 59655133 4.10 2.53% 5.33 90.24% May-00 TX-1316 (A512782, A535113), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 59655135 4.46 2.25% 5.67 94.30% May-00 TX-1316 (EK105973, A582743), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 59655135 7.61 10.80% 5.01 91.63% May-00 TX-1316 (EK105973, A582071), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 39695084 6.10 14.36% 5.18 90.29% Sep-00 TX-1316 (A503428, A617522) 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 59695083 4.71 6.76% 6.17 97.08% Sep-00 TX-1316 (A527475, A617508), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 59780076 5.53 7.18% 5.05 92.64% Jun-01 TX-1316 (A512762, A535114), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 59780076 5.19 3.37% 6.62 94.99% Jun-01 TX-1316 (A512762, A541494), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 79780050 5.83 17.95% 6.37 96.14% Jun-01 TX-1316 (EK111272, A582060), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 79780050 5.31 3.60% 4.83 93.28% Jun-01 TX-1316 (EK111272, A582064), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 59820002 4.18 1.65% 5.92 97.82% Oct-01 TX-1316 (EK105233; 978, A617509), 1368, H HS/LC 1341 (¶¶ 92-4, 121-9) Curix Opthos 59820002 5.35 1.55% 5.17 94.81% Oct-01 TX-1316 (EK105233; 978, A582744), 1368, H LS/HC 1341 (¶¶ 92-4, 121-9) Agfa Opthos 59820002 6.48 6.46% 4.96 97.45% Oct-01 TX-1316 (EK105233; 978, A541493), 1368, H LS/HC 1341 (¶¶ 92-4, 121-9) Curix Opthos 59820004 4.77 8.95% 5.19 88.83% Oct-01 TX-1316 (EK105975, A582747), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 59820004 5.41 4.81% 5.44 97.26% Oct-01 TX-1316 (EK105975, A582070), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 39830019 6.27 1.32% 4.55 89.13% Nov-01 TX-1316 (A503429, A617521), 1368, 1341 H LS/HC (¶¶ 192-4, 121-9) Curix Opthos 59850003 4.80 4.76% 5.71 94.41% Jan-02 TX-1316 (EK105977, A591775, A591792, H HS/LC A617511), 1368, 1341 (¶¶ 92-4, 121-9) Curix Opthos 59850003 6.00 5.67% 4.85 93.86% Jan-02 TX-1316 (EK105977, A582069), 1368, 1341 H LS/HC (¶¶ 192-4, 121-9) Curix Opthos 79850003 4.31 11.89% 5.20 88.46% Jan-02 TX-1316 (EK111271, A582063), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 79850003 6.81 8.64% 4.66 94.88% Jan-02 TX-1316 (A582065, EK111271), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 59850002 5.87 4.30% 6.09 92.81% Jan-02 TX-1316 (A535115, A527474), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 59850002 4.11 1.45% 5.17 90.32% Jan-02 TX-1316 (A527474, A591776, A591793, H HS/LC A617510), 1366, 1341 (¶¶ 92-4, 121-9) Curix Opthos 59850002 7.02 11.88% 4.91 95.68% Jan-02 TX-1316 (A527474, A591777, 94), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 59850002 5.35 1.55% 5.19 94.63% Jan-02 TX-1316 (A582745, A527474), 1366, 1368, H LS/HC 1341 (¶¶ 92-4, 121-9) Curix Opthos 59870075 3.90 1.51% 5.50 92.35% Mar-02 TX-1316 (EK105976, A591774, A591796, 98, H HS/LC A617512), 1368, 1341 (¶¶ 92-4, 121-9) Curix Opthos 59870075 5.12 2.67% 5.23 96.46% Mar-02 TX-1316 (EK105976, A582068), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 59870074 5.42 93.23% Mar-02 TX-1316 (A582746, EK105974), 1368, 1341 H HS/LC (¶¶ 92-4, 121-9) Curix Opthos 59870074 5.55 2.51% 5.12 93.09% Mar-02 TX-1316 (EK105974, A582067), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) Curix Opthos 79890061/01 7.09 5.82% 6.60 97.54% May-02 TX-1316 (A503430, A535127), 1368, 1341 H (total) (¶¶ 92-4, 121-9) Curix Opthos 79890061/01 5.74 14.44% 5.47 87.26% May-02 TX-1316 (A503430, A541483), 1366, 1368, H HS/LC 1341 (¶¶ 92-4, 121-9) Curix Opthos 79890061/01 7.60 10.33% 5.67 96.82% May-02 TX-1316 (A503430, A541482), 1368, 1341 H LS/HC (¶¶ 92-4, 121-9) None of these film samples can be said to fall within the literal scope of claim 1 of the `425 or `520 patents because none literally satisfy the average aspect ratio of greater than 8:1 limitation of the claims.

With regard to claim 1 of the `426 patent, it should be noted that the vast majority of samples did not fall within the scope of claim 1 of the `426 patent. All of the data points except for one obtained from batch 59315075 was well below the "at least 50 percent of the total projected area" requirement of the claim. The one data point seemingly falling within the literal scope range of grain sizes is so isolated relative to the numerous other samples that it cannot be considered a meaningful, reliable data point representative of the product. This data point appears to be an aberration because it is so far removed from the other data points.

It is my recommendation that Opthos H does not infringe literally the asserted claims of the `426, `425, and `520 patents.

Equivalency

A finding of infringement under the doctrine of equivalents requires a showing that the difference between the claimed invention and the accused product was insubstantial. One way of establishing equivalency is to show that the feature found not to be literally included in the accused product is provided by some thing that performs substantially the same function in substantially the same way to achieve the same result. AquaTex Indus., Inc. v. Techniche Solutions, 479 F.3d 1320, 1328 (Fed. Cir. 2007); see also Graver Tank Mfg. Co. v. Linde Air Prods. Co., 339 U.S. 605, 608, 70 S.Ct. 854, 94 L.Ed. 1097 (1950) ("[A] patentee may invoke this doctrine [of equivalents] to proceed against the producer of a device `if it performs substantially the same function in substantially the same way to obtain the same result.'") (quoting Sanitary Refrigerator Co. v. Winters, 280 U.S. 30, 42, 50 S.Ct. 9, 74 L.Ed. 147 (1929)). The function, way, result inquiry focuses on "an examination of the claim and the explanation of it found in the written description of the patent." Vehicular Techs. Corp. v. Titan Wheel Int'l, Inc., 141 F.3d 1084, 1090 (Fed. Cir. 1998); see also Warner-Jenkinson Co. v. Hilton Davis Chem. Co., 520 U.S. 17, 40, 117 S.Ct. 1040, 137 L.Ed.2d 146 (1997) ("An analysis of the role played by each element in the context of the specific patent claim will thus inform the inquiry as to whether a substitute element matches the function, way, and result of the claimed element.").

Kodak, as the party asserting equivalency, was required to prove by a preponderance of the evidence infringement under the doctrine of equivalents. PSC Computer Prods., Inc. v. Foxconn Int'l, Inc., 355 F.3d 1353, 1357 (Fed. Cir. 2004); CVI/Beta Ventures, Inc. v. Tura LP, 112 F.3d 1146, 1161 (Fed. Cir. 1997), cert. denied, 522 U.S. 1109, 118 S.Ct. 1039, 140 L.Ed.2d 105 (1998). Texas Instruments, Inc. v. Cypress Semiconductor Corp., 90 F.3d 1558 (Fed. Cir. 1996), cert. denied, 520 U.S. 1228, 117 S.Ct. 1818, 137 L.Ed.2d 1027 (1997), sets forth the patentee's evidentiary burden under the doctrine of equivalents:

a patentee must . . . provide particularized testimony and linking argument as to the "insubstantiality of the differences" between the claimed invention and the accused device or process, or with respect to the function, way, result test when such evidence is presented to support a finding of infringement under the doctrine of equivalents. Such evidence must be presented on a limitation-by-limitation basis. Generalized testimony as to the overall similarity between the claims and the accused in-fringer's product or process will not suffice.

Id. at 1567. (emphasis added).

Evidence of equivalence must be from the perspective of someone skilled in the art, for example "through testimony of experts or others versed in the technology; by documents, including texts and treatises; and, of course, by the disclosures of the prior art." See Graver Tank, 339 U.S. at 609, 70 S.Ct. 854; see also Texas Instruments, 90 F.3d at 1566. But, while many different forms of evidence may be pertinent, when the patent holder relies on the doctrine of equivalents, as opposed to literal infringement, the difficulties and complexities of the doctrine require that evidence be presented to the fact-finder through the particularized testimony of a person of ordinary skill in the art, typically a qualified expert, who on a limitation-by-limitation basis describes the claim limitations and establishes that those skilled in the art would recognize the equivalents. Centricut, LLC v. Esab Group, Inc., 390 F.3d 1361, 1369-70 (Fed. Cir. 2004), cert. denied, 546 U.S. 814, 126 S.Ct. 337, 163 L.Ed.2d 49 (2005).

In this case Kodak provided insufficient particularized testimony that specifically addressed equivalents "on a limitation-by-limitation basis" or established the insubstantiality of the differences between each limitation of the claims and the accused products. Texas Instruments, 90 F.3d at 1567; see also Lear Siegler v. Sealy Mattress Co. of Mich., 873 F.2d 1422, 1426 (Fed. Cir. 1989) ("It is clear that LSI did not elicit testimony explicitly comparing the claimed and accused devices as to all three elements of equivalence.").

Mr. Dillenbeck was Kodak's primary witness concerning equivalents as to the grain size parameters of the claims of the T-grain patents.⁹ Mr. Dillenbeck's testimony centered on what he termed a "holistic" comparison of the Agfa products to various Kodak products. Trial Tr. 413 (Dillenbeck). The holistic comparison examined film design, emulsion design and product attributes such as speed, contrast, Dmin, Dmax and other properties.

9 In its post-trial submissions Kodak submitted certain appendices apparently attempting to establish the necessary support for equivalency on a limitation-by-limitation basis. While the appendices cited to certain trial exhibits, there was no particularized testimony about them or explanation about how the exhibits supported equivalency. There was no testimony, for example, explaining why a thickness of 0.229 µ was equivalent to the claimed less than 0.2 µ or similar testimony about the other limitations upon which equivalency was based.

This holistic comparison was flawed inasmuch as infringement, either literal or by equivalence, is to be determined by comparing the accused device with the claims in suit, not with a preferred or commercial embodiment of the patentee's claimed invention. Johnson Johnston Assocs. Inc. v. R.E. Serv. Co., 285 F.3d 1046, 1052 (Fed. Cir. 2002) ( en banc); Martin v. Barber, 755 F.2d 1564, 1567 (Fed. Cir. 1985); Gargoyles, Inc. v. United States, 32 Fed. Cl. 157, 161 (Fed. C1. 1994), aff'd, 113 F.3d 1572 (Fed. Cir. 1997). As explained in Amstar Corp. v. Envirotech Corp., 730 F.2d 1476 (Fed. Cir. 1984), cert. denied, 469 U.S. 924, 105 S.Ct. 306, 83 L.Ed.2d 240 (1984), "[i]nfringement is not determined by comparison between commercial products sold by the parties." Amstar, at 1481-82. The identification of the elements of the function, way, result test solely "entails an examination of the claim and the explanation of it found in the written description of the patent," as well as "[i]n some cases[] the patent's prosecution history." Vehicular Techs., 141 F.3d at 1090. This inquiry leaves no room for consideration of the patentee's product. AquaTex Indus., Inc., 479 F.3d at 1328.

Even if a comparison of Agfa products with Kodak products were sufficient for determining infringement under the doctrine of equivalents, Mr. Dillenbeck never made an assessment as to whether any of the Kodak products that he used as comparisons to the accused Agfa products actually fell within the scope of any asserted patent claim. For instance, Mr. Dillenbeck never saw any tabulated data as to grain sizes of any Kodak products and never reviewed any TEMs of any Kodak products. Trial Tr. at 425-27, 454-55 (Dillenbeck).

Moreover, Kodak's arguments that its and Agfa's products are equivalent to one another in function/way/result is not supported by the evidence. At trial, Mr. Dillenbeck rested his position of equivalence primarily on the issue of less silver usage, testifying that the "key advantage with the t-grains is the increased surface area and the ability to have a lower amount of silver that is utilized." Trial Tr. at 341-42, 337-38 (Dillenbeck). None of the claims specifically address silver usage, however. Additionally, the `425 and `426 patents touted the advantage of crossover reduction achievable by providing an X-ray film with a spectrally sensitized tabular grain emulsion having the specific selection of grain size parameters set forth in the claims. None of the claims specifically address crossover, however.

Even if relevant, a review the silver coating weights and screen light crossover as measured by Kodak for the accused Agfa products as compared to the Kodak products that Mr. Dillenbeck was familiar with and with which he made his comparisons establishes that Agfa's products use more silver and have higher crossover as seen in the following table: Silver Coverage Trade Name (mg/ft² ) % Crossover TX No. Cronex 10T (578) Cronex 10TL (579) Ultravision C (565) Ultravision Ci (560) Ultravision L (564) Ultravision G Orthovision G (525) Orthovision L (527) Agfa B Plus (CX1X5)* Curix Ortho HT-G (CXODGUV2)* DI Image Plus Green (CXODGGV2)* Kodak TMAT G/RA Kodak Xomat K (nontabular) Agfa OPTHOS H (CXODGAS)* Kodak INSIGHT

172, 213/175 28.47, 12.38 562, 567(EK115219, 224), 1262, 1264(EK111738), 1290(EK112130), 454(EK063893), 455(EK063951), 1262(EK049789) 156-157 24.45, 9.48 590, 1264(EK111738), 454(EK063893) 203, 206 8.01 90(EK049603), 572, 490(EK111805), 378, 454(EK063893), 455(EK063957, 969) 156, 166 10.30 568(EK115157), 490, 378, 454(EK063893), 455(EK063969) 207, 199 10.73 30.88, 390(EK049604), 570, 378, 454(EK063893), 12.36 455(EK063957), 1264(EK111738) 199, 200, 201, 31.37, 11.89 455(EK063944, 964), 390(EK049602), 204 567(EK115260), 1264(EK111738), 1262(EK049789) 193 588, 378 179 567(EK115237), 586, 378 177, 187, 176 35.99, 37.06 574, 578, 576, 377 437, 520(EK112173), 1283(EK111844) 225, 212.4, 205, 27 390(EK049601), 580, 581, 377, 1274, 1364, 194 454(EK063893), 455(EK063925) 182-194 29 1364, 1214 148 18.3 390(EK049600), 490(EK111805), 455(EK063925), 1262(EK049789) 206 31.37 1262, 1283(EK111844), 1262(EK049789) 239/237, 226/215, 390(EK049599), 567(EK115253), 377, 224/229, 233 454(EK063893), 455(EK063937) 178/197 390(EK049599) See also Trial Tr. at 376-93 (Dillenbeck), 971-75 (Van den Zegel), 1066-69 (Hudson).

The accused Agfa products have higher silver coating weights and higher screen light crossover than the Kodak products to which they were compared. The coating weights and crossover values for the Agfa films are also in line with coating weights and crossover values for Kodak's nontabular blue radiographic film. Trial Tr. at 1067-68 (Hudson).

With regard to the function/way/result test, Kodak argues that the function of the grain dimension limitations in the patents was to "provid[e] grains with development centers which, when exposed to light, form an invisible latent image, which image can then be turned into a visible image via use of a chemical developer." Kodak FOF No. 109. Kodak asserts that Agfa's emulsion performs this function in the same way, i.e., "through use of flat, thin grains, which enable this functionality at significantly lower cost than non-T-grain emulsions." Id. The substantially same result is, according to Kodak, "an image suitable for diagnostic purposes." Id. Kodak does not have a claim as broad as it asserts under equivalency and did not seek such a broad claim. Indeed, it admitted in the patents that high aspect tabular grains were prior art and that No-Screen was a tabular grain product. Kodak's function/way/result analysis would potentially impermissibly cover this admitted prior art.

Mr. Dillenbeck implicitly testified that the accused products must perform substantially the same function as the Kodak products in substantially the same way to achieve substantially the same result because Agfa's products competed with Kodak's. The conclusions are clearly directed to the products as a whole and not to the claimed limitations, however. Generalized testimony as to overall similarity between the claims and the accused infringer's product cannot support a finding that the differences are insubstantial. Texas Instruments, 90 F.3d at 1568. Similarly, various exhibits which Kodak pointed to in its Post-Trial Findings of Fact do not provide a limitation-by-limitation analysis or explanation between the claims and the accused products.

In certain instances, Kodak asserts equivalency by increasing the size of grains necessary to achieve the specified average total area. There was no evidence establishing why the increased size of the grains necessary to achieve that area was an insubstantial difference Expanding the grain size limitations would begin to call into question the differences with the state of the prior art. Kodak did not use the modifier "about" when setting the thickness limitation, for example, as is common in patent prosecution.

Kodak has not delineated where the range of equivalents purportedly extends. Kodak's asserted equivalence differs within and between different products. With regard to the `426 patent for example, Kodak asserts scopes of equivalency of 0.236 µ, 0.235 µ, and 0.227 µ for Ultravision C. For Ultravision L, the scope of equivalency has been stretched to 0.246 µ and 0.259 µ. For Agfa CP-BU, Kodak asserts equivalency of 0.229 µ, 0.226 µ, 0.227 µ, 0.229 µ, and 0.230 µ. For Curix Ortho HT-G, the scope of equivalence fluctuates between 0. 208 µ, and 0.255 µ. The only trait that these equivalent thicknesses apparently share is that in each sample the thickness was selected to obtain an average aspect ratio and projected area close to or within the claim limitations. The scope or range of equivalency is not determined by examining the accused product, however.

While Kodak is entitled to a range of equivalence for the grain size limitations, for example, more proof is required than an assertion that the products compete in the market.

For these reasons, it is my recommendation that Kodak be found not to have proven infringement of any of the accused products under the doctrine of equivalents.

Unenforceability

Agfa asserts that Kodak's failure to disclose the Tani articles and its representations regarding the materiality of the Tani IX publication were intentional and misleading. Trial Tr. at 1126-29 (Liggero). Agfa asserts that the Tani series was particularly material because it discloses optimally chemically and spectrally sensitized tabular grains, which fills the gap Kodak said was lacking in the prior art. Trial Tr. at 1128-29, 1187 (Liggero).

Inequitable conduct entails a two-step analysis: "first, a determination of whether the withheld reference meets a threshold level of materiality and intent to mislead, and second, a weighing of the materiality and intent in light of all the circumstances to determine whether the applicant's conduct is so culpable that the patent should be held unenforceable." GFI, Inc. v. Franklin Corp., 265 F.3d 1268, 1273 (Fed. Cir. 2001). Unenforceability due to inequitable conduct must be established by clear and convincing evidence. Gross negligence is not enough. Kingsdown Medical Consultants v. Hollister Inc., 863 F.2d 867, 876 (Fed. Cir. 1988) (en banc), cert. denied, 490 U.S. 1067, 109 S.Ct. 2068, 104 L.Ed.2d 633 (1989).

Agfa points out that Kodak never disclosed the Tani articles during the examination of the applications that resulted in the `425 and `426 patents. TX 1005-07; Trial Tr. at 1126-27 (Liggero). Agfa also alleges that Kodak never disclosed Tani III, V or VII during the examination of the application that resulted in the `520 patent, and never submitted Tani IX via an information disclosure statement during examination. TX 1005-07, 1009, 553 and 556; Trial Tr. at 1126-29 (Liggero). Kodak did submit Tani IX in an information disclosure statement after allowance and before issuance, but was unsuccessful in having it considered on the record. TX 1321 (EK0004931-61).

Dr. Booms testified that although he was sure he had at some time read the Tani articles, he could not recall precisely when he had read them. Rather surprisingly, Agfa failed to question Dr. Booms about the significance of the Tani articles and their applicability to the subject matter of the claims of the `520 patent. At least one of the Tani articles is listed as a considered reference on the face of the `520 patent. Despite its submission, no rejection was made based upon it. Based on the examiner's indifference to the cited Tani article and for the reasons expressed above in the invalidity analysis section, the intent and materiality of the series are neither clear nor convincing.

Agfa asserts that the `520 patent is unenforceable due to inequitable conduct surrounding Kodak's alleged withholding of the Wheat memorandum, an internal Kodak document. TX 1221. Kodak made statements to the PTO to the effect that the `520 patent disclosed the "most pertinent" prior art. Trial Tr. at 1132 (Liggero). Kodak also stated that the comparative examples of its `520 patent were "optimally chemically and spectrally sensitized." According to Agfa, these statements were demonstrably false in view of the material described in the Wheat memorandum which was superior to and more optimally sensitized than the comparative examples disclosed in the `520 patent (5-7 of Figure 5). TX 1334 (¶¶ 65-71), 1221; Trial Tr. at 1130-33 (Liggero).

Agfa is unable to meet its burden with regard to the threshold level of materiality. In Figure 1 in the Wheat Memorandum, the group consisting of inventive emulsions "4-7" exhibits much better speed-granularity ratios than does the group consisting of emulsions "1-3." TX 1221, 835 (10/26/06 Marchetti report, ¶ 50); Trial Tr. 122 (Marchetti) (Wheat Memorandum shows "a paradigm shift that demonstrates the invention"), 114 (Booms); Depo. Desig. Tab 36 (06/06/03 Roberts) at 374 (according to the graph, examples 4-7 all show an improvement over 1-3). Emulsion "5" exhibits a materially better speed-granularity ratio than emulsion "3." TX 835 (10/26/06 Marchetti report), ¶ 50. While the emulsions "5" and "3" of the Wheat Memorandum appear to have similar speeds, inventive emulsion "5" has a significantly better granularity position than control example 3. Id. Because the inventive examples show an unexpected improvement over the comparative examples of the Wheat Memorandum, there is no reason to conclude that including control emulsions 1-3 of the Wheat Memorandum in the `520 patent as comparative examples would have had any impact on the prosecution of the `520 patent.

Agfa is also unable to meet its burden of proof with respect to its allegation that Kodak misrepresented its comparative examples as optimally chemically and spectrally sensitized. Agfa bases its allegation on the addition of thiocyanate to the comparative examples of the Wheat Memorandum, whereas the comparative examples of the `520 patent were chemically sensitized in the absence of thiocyanate. TX 1334 (¶¶ 65-71), 1221; Trial Tr. at 1130-32 (Liggero). According to Agfa, the use of thiocyanate during chemical sensitization of the comparative examples of the Wheat memorandum is responsible for material differences between the speed-grain position of the comparative examples of the Wheat memorandum and those of the `520 patent. TX 1334 (¶¶ 65-71), 1221; Trial Tr. at 1130-32 (Liggero).

The testimony established that the precise effect of thiocyanate as either beneficial or detrimental to sensitization in any particular emulsion is unknown, and there is insufficient evidence to conclude clearly and convincingly that the omission of thiocyanate from the patent's comparative examples necessarily reduced the speed/granularity ratio. Trial Tr. 112 (Booms). Similarly, the record establishes that it is possible to apply a different set of sensitization conditions to two emulsions — Seven two identically precipitated emulsions — and reach a similar speed-granularity ratio. Trial Tr. 1229-32 (Marchetti); TX 835 (10/26/06 Marchetti report), ¶ 48; Depo. Desig. Tab 32 (02/14/07 Liggero) at 305-06.

Agfa is also unable to meet its burden with regard to intent to deceive. There is no clear and convincing evidence of any misrepresentation by any Kodak representative to the PTO, much less any evidence of any material misrepresentation in connection with the prosecution of the `520 patent. There is no clear and convincing evidence that any Kodak representative at any time harbored any intent to deceive the PTO. Agfa Depo. Desig. Tab 16 (09/22/04 Thomas) at 96; Trial Tr. 1131 (Liggero) ("could have been an oversight").

Agfa argues that the inventive examples of the Wheat memorandum underlie the examples of the `520 patent. The Liggero Report concludes that Emulsions 3 and 4 in Fig. 5 of the `520 patent are in fact emulsions "4" and "5" of the Wheat Memorandum. Agfa argues that inventive examples from the Wheat memorandum were copied to the `520 patent and that the comparative examples of the Wheat memorandum were omitted intentionally. This according to Agfa is evidence of intent to deceive.

None of the fact witnesses who testified regarding the Wheat Memorandum re- called whether the examples in the Wheat Memorandum were the same set of examples as in the `520 patent. Trial Tr. 106-07 (Booms); Agfa Depo. Desig. Tab 16 (09/22/04 Thomas) at 85-86, 87, 94; Agfa Depo. Desig. Tab 22 (11/05/04 Wheat) at 43-45, 68. The fact that both sets use the same number of examples has no bearing on the identity of the examples because it was "a pretty standard way" to prepare comparative examples using at least three inventive examples and at least three control examples. Trial Tr. 106-07 (Booms). The fact that some of the emulsion preparation factors (iodide content and average aspect ratio) were the same is not clear and convincing proof of their identity because many emulsions were made that had similar properties, but it is still "very hard" to tell if they are the same emulsions. Agfa Depo. Desig. Tab 16 (09/22/04 Thomas) at 71, 87; Agfa Depo. Desig. Tab 22 (11/05/04 Wheat) at 44.

Agfa's arguments about the Wheat Memorandum fail to clearly and convincingly establish inequitable conduct.

Agfa argues also that Kodak's distinction of its own `520 patent during the reexamination proceedings to confirm patentability of the `426 patent claims was intentional and misleading. Trial Tr. at 1134 (Liggero).

During reexamination of the `426 patent, the PTO rejected claims over the `520 patent. Kodak overcame the rejection by stating that the `520 patent "is entirely irrelevant to the claimed invention. It contains no mention of any radiographic element having a two-sided format. It is devoid of any teaching directed to the problem of crossover." TX 1007(50); Trial Tr. at 1133-34 (Liggero).

These statements were incorrect. The `520 patent teaches the use of a radiographic element — "[w]hen the photographic elements of the invention are intended to serve radiographic applications" — and incorporates by reference Research Disclosure, Item 18431. TX 312 (520 patent, col. 41, ll. 31-36); Trial Tr. at 1133-34 (Liggero). Research Disclosure, Item 18431 describes the problem of crossover in a radiographic element. TX 1036; Trial Tr. at 1134 (Liggero).

Kodak admits that this statement was "mistaken," but argues that it was immaterial because the PTO had the `520 patent before it and could readily determine what disclosures were made, or not made, in the `520 patent's specification and claims. Kodak also argues that the misstatement concerning the `520 patent was harmless because its disclosure of a radio-graphic element was cumulative of other references that were cited and discussed during the reexamination.

There is insufficient evidence that any Kodak representative at any time clearly and convincingly harbored any intent to deceive the PTO. Mr. Thomas testified that he made the statement based on his memory at the time of the `520 patent without checking because he felt confident in his memory at the time — demonstrating that there was no intent to mislead. Agfa Depo. Desig. Tab 17 (09/23/04 Thomas) at 226. While Mr. Thomas was possibly negligent in not checking the `520 patent before making the statement, more than gross negligence is required for inequitable conduct. Kingsdown Medical Consultants, 863 F.2d at 876.

It is my recommendation that none of the T-grain patents in suit be found unenforceable.

Willfulness

There can be no willful infringement where there is no infringement. In view of my recommendation that Agfa has not infringed the claims of the patents in suit with the exception of infringement of the `426 patent by Agfa's Orthovision G, either literally or under the doctrine of equivalents, Kodak's willful infringement claim is not viable.¹⁰

10 While the Cronex 10TL product was also found to infringe, Kodak cannot recover damages because of lack of notice under 35 U.S.C. § 287.

Even if Kodak had established infringement of any or all of Agfa's products, Kodak has not established willful infringement by clear and convincing evidence. Comark Commc'ns, Inc. v. Harris Corp., 156 F.3d 1182, 1190 (Fed. Cir. 1998) (the patentee bears the burden of persuasion and must prove willful infringement by clear and convincing evidence); Amsted Indus. Inc. v. Buckeye Steel Castings Co., 24 F.3d 178, 181 (Fed. Cir. 1994).

Willfulness is assessed at the time the infringer received notice of the existence of the patent. Johns Hopkins Univ. v. CellPro, Inc., 152 F.3d 1342, 1362 (Fed. Cir. 1998). Here, that is in 1995 for Cronex 10T and April, 2001, for the remaining products except for Cronex 10TL for which no timely notice was given.¹¹

11 Indeed, Mr. Waaser's November 1995 letter to Kodak forwarding the Cronex 10T production codes stated that he was confident Kodak would agree that the patents were not infringed after it conducted its further evaluation. Kodak did test subsequent Cronex 10T specimens and did not until 2001 accuse the product of infringement. Defendants cannot be considered to be acting recklessly where Kodak had every opportunity to accuse the product of infringement and yet remained silent.

While at one time a potential infringer, with actual notice of another's patent rights, had an affirmative duty to exercise due care to determine whether or not it is infringing, nCube Corp. v. Seachange Int'l, Inc., 436 F.3d 1317, 1324 (Fed. Cir. 2006) ("Actual notice of another's patent rights triggers an affirmative duty of due care to avoid infringement."); Underwater Devices, Inc. v. Morrison-Knudsen Co., 717 F.2d 1380, 1389-90 (Fed. Cir. 1983), overruled in part on other grounds by Knorr-Bremse Systeme Fuer Nutzfahrzeuge GmbH v. Dana Corp., 383 F.3d 1337 (Fed. Cir. 2004) ( en banc), the Federal Circuit in In re Seagate Technology, LLC, 497 F.3d 1360 (2007) overruled those prior holdings. Now, a finding of willfulness "requires at least a showing of objective recklessness." Seagate Technology, LLC, 497 F.3d at 1371. A "patentee must show by clear and convincing evidence that the infringer acted despite an objective high likelihood that its actions constituted infringement of a valid patent." Id.

While the Federal Circuit in Seagate Technology, LLC, "le[ft] it to future cases to further develop the application of the standard", id., it did not reject using factors under the totality of the circumstances. Liquid Dynamics Corp. v. Vaughan Co., 449 F.3d 1209, 1225 (Fed. Cir. 2006); Imonex Servs., Inc. v. W.H. Munzprufer Dietmar Trenner GMBH, 408 F.3d 1374, 1377 (Fed. Cir. 2005).

"Willful infringement is not established by the simple fact of infringement," even where the accused has knowledge of the patents. Id.; accord Ajinomoto Co. v. Archer-Daniels-Midland Co., 228 F.3d 1338, 1351-52 (Fed. Cir. 2000). "The patentee must present threshold evidence of culpable behavior" before the burden of production shifts to the accused to put on evidence that it acted with due care. Norian, 363 F.3d at 1332 ("Absent an initial presentation of evidence . . . this burden of coming forward in defense [does] not arise."). That threshold showing cannot be satisfied merely by proof that the accused is asserting the attorney-client privilege to withhold an opinion of counsel. As explained in Knorr-Bremse, when the attorney-client privilege and/or work product privilege is invoked by a defendant in an infringement suit, it is not appropriate for the trier of fact to draw an adverse inference with respect to willful infringement. Knorr-Bremse Systeme Fuer Nutzfahrzeuge GmbH v. Dana Corp., 383 F.3d 1337, 1344 (Fed. Cir. 2004) ( en banc).

DuPont, Sterling, Agfa-Gevaert, and Agfa Corporation were aware of the Kodak patents and took the patents into account when they deliberately designed their products to avoid infringing the Kodak patents. Mr. Hawley testified to that effect with regard to the 1995 meeting between Kodak and DuPont. Dr. Hudson testified that the metes and bounds of the patents were known and taken into account when designing film products:

Q. And what was your understanding or did you have an understanding of the scope of those patents as a scientist working in this field?

A. Yes, I did. I mean, my understanding was that there are thickness requirements, there are aspect ratio requirements and total percent projected area of the total projected area requirements for these patents, and they vary a little bit in the details.

. . .

Q. And more specifically, do you recall any directions or instructions that you were provided insofar as how to avoid the Kodak patents?

A. Yes, we were supposed to stay above .2 or and/or an aspect ratio less than 5:1 and/or the total percent projected area of the grains less than the 50 percent.

Trial Tr. at 802-3 (Hudson), Dr. Hudson testified that "[i]t was DuPont's policy to stay out of all, all of our competitor's patents or anyone's patents." Trial Tr. at 802 (Hudson). He testified further that DuPont, Sterling and Agfa designed products to avoid the Kodak patents:

Q. Let me ask you: Did you believe you were entitled to design and manufacture films that were outside the limits of those patents?

A. Yes, I mean for one, Kodak did not patent any film with a thickness less than .2 because there are the additional requirements, there are aspect ratio requirements.

Q. Did you at Dupont, Sterling or Agfa ever design or manufacture a film that you knew fell within the scope of the Kodak tabular grain patents?

A. No.

Q. Did you at DuPont, Sterling or Agfa ever sell, do you know of any selling of any product that had grains that fell within the scope of the Kodak tabular grain products?

A. No.

Trial Tr. at 862 (Hudson). Dr. Van den Zegel testified that Agfa designed their products to be "out of" the grain size parameters of the Kodak patents. Trial Tr. at 958 (Van den Zegel); see also TX 48; Trial Tr. 802-03, 862 (Hudson).

DuPont, Sterling, and Agfa were conscious of the need not to infringe and made a bona fide effort to avoid infringement by attempting to design around the claimed invention. Rolls-Royce Ltd. v. GTE Valeron Corp., 800 F.2d 1101, 1109-10 (Fed. Cir. 1986). DuPont, Sterling, and Agfa proceeded in good faith to avoid the patents. American Original Corp. v. Jenkins Food Corp., 774 F.2d 459, 465 (Fed. Cir. 1985). They expended considerable effort in testing emulsions using EGSA or Moeller counter instruments.

Q. Now, all right. Let's go on. And on what basis do you say did they fall outside of the scope of the tabular grain patents by Kodak?

A. Via our testing of the EGSA on every batch of grains that we produced at Brevard, which is a 5 to 7,000 grain measurement on every batch.

Trial Tr. at 824 (Hudson). Dr. Hudson testified that the EGSA method is consistent with the average approach used in the tabular grain patents:

Q. Is the EGSA method that you described consistent with the suggestion to use an averaged approach or to use an approach that involves averages?

A. Yes. The EGSA approach is an average approach that is suggested by this passage.

Q. And why do you say that?

A. Because they allow you to use an average value and then I'm going to get an average value for the thickness from the SEM, and also, because the EGSA method is going to average 5 to 7,000 grains.

Trial Tr. at 807-8 (Hudson). Dr. Hudson testified that a Moeller counter is the same as an EGSA:

A Moeller counter is the same device that I described earlier as an EGSA. In other words, Dr. Moeller presented his idea in the early 70s and then the various manufactures kind of developed their own EGSA or Moeller counter device. At Mortsel within Agfa in their Moeller counter they count 10,000 grains, where generally for the tabular grains that we made at Sterling that eventually became Agfa-Gevaert, we generally count the 5 to 7,000 grains. They display their information a little bit differently, but the underlying principle is exactly the same.

Trial Tr. at 1060-61 (Hudson). Dr. Van den Zegel testified that the Moeller counter was used at Agfa to measure average grain volume for every production emulsion:

Q. What is it a Moeller measurement?

A. Moeller measurement measures, you have the silver halide and then measured with an electrode how much electricity is needed to reduce the silver plus to silver, and that amount of electricity is a measure for the volume of the crystal and then many crystals are measured, about 10,000, and a mean volume and a mean diameter is calculated.

Q. And are those parameters tracked for production?

A. Yes.

Q. And how are they tracked?

A. They are measured for each emulsion and they bring into a database.

Trial Tr. at 966 (Van den Zegel); see also TX 1351.

Kodak correctly points out that these measurement techniques do not enable determination of grain thickness and diameter on a grain-by-grain basis. Depo. Desig. Tab 21 (02/05/04 Hudson) at 19. See Trial Tr. 865 (Hudson); Trial Tr. 909 (Hudson); Trial Tr. 192-93 (Cole); Depo. Desig. Tab 21 (02/05/04 Hudson) at 18-19.

While the techniques employed by Agfa for monitoring products were not without their flaws, the shadowed electron microscopy technique identified in the T-grain patents had its own flaws as discussed above. The TEM method identified in the patents to measure the grains is impractical on a commercial scale, as evidenced by the fact that neither Kodak nor Agfa use TEMs to monitor their products during manufacturing. TX 1341 (¶ 43); Trial Tr. at 964-65 (Van den Zegel). The grain volume monitoring equipment (EGSA and Moeller counter) measured and provided average grain volume data for thousands of grains for each batch sample, Trial Tr. at 803-04 (Hudson), 966, 1060-61 (Van den Zegel), whereas the shadowed electron microscope technique samples about 100-150 grains. The potential for error created by the sampling and measurements accompanying shadowed electron microscopy is explained by Dr. Apley and Kodak acknowledged that Mr. Van Dam's accuracy was ± 0.022 µ. Under the circumstances, use of EGSA and the Moeller counter cannot be viewed as objectively reckless.

While there was testimony that DuPont had cleared the DuPont/Sterling products as part of the product review, no opinions to that effect were introduced. Likewise, while Agfa presented testimony that its patent agents had cleared the products, no opinions were introduced and any such opinion would be of little relevance because a patent agent, particularly a foreign patent agent, is not qualified to opine on infringement of a U.S. patent. Spindelfabrik Suessen-Schurr v. Schubert Salzer Maschinenfabrik AG, 829 F.2d 1075, 1084 (Fed. Cir. 1987), cert denied, 484 U.S. 1063, 108 S.Ct. 1022, 98 L.Ed.2d 987 (1988).

Despite the failure to offer exculpatory opinions, the record established that DuPont, Sterling and the Agfa defendants were aware of the patents, and took steps that they believed avoided the patents. Further, the Defendants were not advised about infringement concerns until 2001. Considering that DuPont/Sterling and Agfa openly sold their products and received no accusations of infringement between 1995 and 2001 despite Kodak's testing of at least Cronex 10T, it cannot be said that they proceeded objectively recklessly in disregard of the patents. It was not unreasonable for Defendants to believe that products in the market for years were free of infringement. Kodak, after all, had a group charged with monitoring the products of its competitors. Recall that Kodak notified DuPont in 1995 about Cronex 10T, the parties met, Kodak sampled the product, and it was not until 2001 that Kodak again claimed Cronex 10T as then manufactured infringed.

Moreover, as noted above, the record established that substantial questions were raised at trial as to validity and infringement. Seagate Technology, LLC, at 1378.

In view of the above, it is recommended that it be found that Kodak has failed to sustain its clear and convincing burden of proof on willfulness.

Conclusion

In view of the above, it is recommended that Cronex 10TL and Orthovision G be found to infringe literally, that the other accused products be found to not infringe literally or by equivalence, that Kodak did not provide notice regarding Cronex 10TL prior to expiration of the T-grain patents, that Defendants have not established that the asserted claims are invalid, that Defendants have not established that the patents are unenforceable, and that Kodak has not established that any infringement was willful.

The parties are reminded of the thirty-day time period and page limit limitations applicable to any objection that they may have as to this Report and Recommendation.

Sept. 27, 2006.

REPORT AND RECOMMENDATION ON EVIDENTIARY ISSUES

JOSEPH W. BERENATO, III, Special Master.

INTRODUCTION

Defendants Agfa-Gevaert N.V. and Agfa Corporation (collectively "Agfa") have requested reconsideration of my trial decision refusing admission of Trial Exhibit (TX) 1255. Eastman Kodak Company ("Kodak") had objected to admission of TX 1255 on hearsay and other grounds. During trial, Kodak's objection was sustained, but Agfa was given the opportunity to submit a memorandum of law in support of their position that Kodak's objection should be overruled and TX 1255 admitted into evidence. Trial Tr. 579-80. Agfa submitted its memorandum on June 20, 2007. Kodak in turn requested an opportunity to respond to the memorandum in the event that I reconsidered my decision to sustain Kodak's objection and exclude TX 1255. See June 20, 2007 letter from Richard Rochford. Because Kodak's objections are established on the record, there is no need for a responsive memorandum from Kodak.

Additionally, Agfa has objected to the admissibility of certain deposition testimony designated by Kodak pursuant to the Pre-Trial Order. In particular, Agfa asserts that because Mr. Waaser and Dr. Hudson were non-parties to the case and testified at trial, their respective fact deposition testimonies are not admissible.

FACTS

TX 1255 is a four-page handwritten document that Kodak produced to Agfa in response to one of Agfa's requests for production of documents. The document was marked by Kodak with Kodak production numbers EK063372-375, and was marked "Confidential C.A. 02-CV-6465T" prior to its production. Kodak also referred to the exhibit indirectly in support of an interrogatory request, identifying the document as relevant to damages.

Agfa argues that TX 1255 is a memorandum containing minutes taken at a November 1995 meeting between representatives of Kodak and E.I. du Pont de Nemours Company ("DuPont"), predecessor in interest to Defendants. The meeting took place in response to Kodak's notice to DuPont of alleged infringement of DuPont's Cronex 10T product by a letter dated September 7, 1995 (TX 1087). Although the author of the memorandum has not been identified, Agfa presumes that the memorandum must have been authored by one of two members of the Kodak team who are now deceased, given that none of the surviving attendees has claimed authorship. There was no testimony that anyone was charged with taking notes or that anyone recalled anyone taking notes.

Germane to this discovery dispute is the claim language of the patents discussed at the November 1995 meeting. One of those patents, U.S. Patent No. 4,425,426 ("the `426 patent") recites a radiographic element having as an improvement

first imaging means containing tabular silver halide grains having a thickness of less than 0.2 micron and an average aspect ratio of from 5:1 to 8:1 accounting for at least 50 percent of the total projected area of said silver halide grains present in said silver halide emulsion.

The first and second pages (EK063372-73) of TX 1255 seem to address the counting of silver halide grains, stating in part as follows: sic sic DuPont Take total projected area of 25 grains Kodak Take total projected area of 25 grains

Question arose about counting grains: what do you do about the grain[s] 0.2 µ? 1. Agreed to how to measure grains. That only 0.2 µ cout []. DuPont: couts [] all grains less than 0.2 µ Then of those 0.2 µ grains counts only those that have aspect ratio of 5-8:1 to the calculate % projected area. 100 grains 50 0.2 µm 25 .2 µ. and aspect ratio of 5:1 8.1 = % projected area Take total projected area of 100 grains = % projected area Take total projected area of 50 grains DuPont suggest counting all grains Kodak counts only those which are .2µ According to Agfa, the written statement in TX 1255 that "only0.2 cout [ sic]" is critical to Agfa's defenses of inadequate notice under 35 U.S.C. § 287(a) and laches, and Kodak's construction of the `426 patent claims under the doctrine of equivalents. Agfa's Statement of Facts (SOF) No. 51, reproduced in part below, explains the criticality of the written statement:

Kodak never suggested to DuPont at the meeting or thereafter that grains thicker than 0.2 micron should be considered in performing an infringement analysis of the `426 patent under the doctrine of equivalents. Notes taken during that meeting state: "Questions arose about counting grains: What do you do about grain 0.2µ? 1. Agreed to how to measure grains. That only 0.2 cout [ sic — count]." TX 1255

ANALYSIS

1. TX 1255

TX 1255, having been offered to prove the truth of the matters asserted in it, is hearsay. See Fed.R.Evid. 801(c). Agfa does not appear to dispute this conclusion. Rather, Agfa argues that the document is nonetheless admissible under either of the following hearsay exceptions: (i) as an admission against interest under Fed.R.Evid. 804(b)(3); or (ii) as a business record kept in the ordinary course of business under Fed.R.Evid. 803(6).

Fed.R.Evid. 804(b)(3)

TX 1255 does not contain a statement against interest under Fed.R.Evid. 804(b)(3). Agfa would have me read the statement "only 0.2 cout [ sic]" to constitute an admission by Kodak that an agreement was reached between the parties that tabular silver halide grains that are 0.2 micron or greater in thickness do not infringe the patent claims.

According to the testimony at trial, there was no agreement reached at the 1995 meeting regarding "particular production parameters to be practiced" by DuPont (Answer to the Amended Complaint and Counterclaims of Agfa Corp., Fifteenth Affirmative Defense) or regarding the interpretations of the claims of Kodak's patents. Trial Tr. 509 (Hawley).

Bernard Apple, a DuPont representative who attended the meeting, testified in his deposition as follows:

Q. Was there any agreement that if DuPont kept the thickness at a specific number that that would not be infringing?

A. Come again? If they kept . . .

Q. If they maintained a specific thickness for the grains, that would not be infringing. Was there any agreement of that detail?

A. I don't believe so.

Depo. Desig. Tab 2 (01/29/04 Apple) at 206-07. Remus Hudson, another DuPont attendee of the meeting, testified as follows:

Q. Did — at any time at that — during that meeting, did the people from Kodak agree that an emulsion having any particular average grain thickness would not infringe the Kodak patents?

A. No, not that I recall.

Depo. Desig. Tab 21 (02/05/04 Hudson) at 139.

Mr. Hawley, a patent attorney for Kodak who attended the meeting, testified that "there were no agreements at this meeting with regard to anything, really." Trial Tr. 547 (Hawley). Dr. Hudson at trial unequivocally agreed: "I don't think there was any agreement reached between the parties." Trial Tr. 860 (Hudson). See also Trial Tr. 159 (Van Dam) ("Q: And is it in fact true that there was agreement reached as to how to measure the grains, that only grains less than .2 would count; isn't that true? A. No, not that I remember.").

There is no indication that the statements in the memo were made in the context of equivalency discussions. To the contrary, trial testimony showed that the doctrine of equivalents was not discussed at the meeting.

Further, there is no indication from whom the statement originated. Even if the statement had originated from a Kodak employee, it can hardly be considered a "statement which was at the time of its making so far contrary to the declarant's pecuniary and proprietary interest . . . that a reasonable person in the declarant's position would not have made the statement unless believing it true," as required by Rule 804(b)(3).

The statement in question is hardly a statement against interest in the manner asserted by Agfa. Agfa reads the excerpt of TX 1255 in a vacuum, isolated from its context in the document as a whole. The excerpt seemingly relates to a disagreement between Kodak and DuPont at the 1995 meeting as to how to calculate percent projected area under the patent claims.

At the 1995 meeting, there was a disagreement between Kodak and DuPont as to how to calculate the percent projected area. Although recollections at trial regarding the particulars of the disagreement were hazy, the general consensus was that the disagreement involved which grains to include in the numerator of the equation for determining percent projected area. Trial Tr. at 532-35 (Hawley), 580-83 (Waaser), 858-62 (Hudson). Taken in context, the statements at issue regarding the 0.2 micron thickness limitation possibly related to differing literal claim constructions between Kodak and DuPont, not to whether the claims were entitled to equivalence of greater than 0.2 micron.

The reliability and trustworthiness of TX 1255 is also deeply in question. Agfa, the party seeking admittance of the exhibit, represented in its Statement of Fact No. 49 that:

After the presentations, Kodak and DuPont recognized that they were not following the same infringement analysis and that this was one reason why they were coming to different results. Kodak was measuring all grains, segregating out the ones that were less than 0.2 microns, calculating the average aspect ratio of this subset of grains to determine if it was between 5:1 and 8:1, and then, if so, determining whether this subset of grains represented at least 50% of the total surface of all grains calculated. DuPont was measuring all grains, segregating out the grains that were less than 0.2 microns, further segregating out the grains that were thinner than 0.2 microns and had an aspect ratio of between 5:1 and 8:1 and then determining whether this subset had a surface area that was at least 50% of the total surface area of all grains.

The description of the disagreement in TX 1255 is inconsistent with Agfa's characterization of the disagreement. The author of TX 1255 appears to set forth an example describing Kodak's interpretation of percent projected area as the total projected area of 25 grains (having a thickness less than 0.2 micron and an aspect ratio of 5:1 to 8:1) divided by the total projected area of 50 grains (having a thickness of less than 0.2 micron). Further down the page the author states that while "DuPont suggests counting all grains," "Kodak counts only those which are .2, µ" an apparent reference to the denominator in Kodak's calculation. By Agfa's own admission (Agfa's SOF No. 49), this is not an accurate portrayal of the substance of the disagreement. If Agfa's characterization of Kodak's position were correct, Kodak had presented a definition that would have calculated percent projected area as the total projected area of 50 grains (having a thickness of less than 0.2 micron) divided by the total projected area of 100 grains ( i.e., the total population). See Trial Tr. 534 (Hawley).

There are other issues that cast further doubt over the reliability of TX 1255. Several key notations in the trial exhibit, including the statement that "only 0.2 cout [ sic]" have been scratched through and written over, raising uncertainties as to by whom and when these revisions were made. It appears also that TX 1255 may include the handwriting of multiple persons, and there are annotations that possibly were made after the fact. In short, there are too many unanswered questions about this document to support admission of TX 1255 based upon the admission against interest exception to the hearsay rule. After all, considering that representatives at the meeting testified that no agreements were reached at the meeting, who made the agreement that represents the admission?

For these reasons, TX 1255 is not admissible under the admission against interest hearsay exception.

Fed.R.Evid. 803(6)

The Second Circuit has adopted a "generous view" of the business records exception, "construing it to favor [] the admission of evidence . . . if it has any probative value at all." United States v. Freidin, 849 F.2d 716, 722 (2d Cir. 1988) (internal quotations omitted). Although the "principal precondition" to admissibility is the sufficient trustworthiness of the record, Saks Int'l, Inc. v. M/V Export Champion, 817 F.2d 1011, 1013 (2d Cir. 1987), the proffered record must meet all of the requirements of the exception. Admissibility under Rule 803(6) requires both that a memorandum have been "kept in the course of a regularly conducted business activity" and also that it was the "regular practice of that business activity to make the memorandum. . . ." Moreover, all this must be shown by the "testimony of the custodian or other qualified witness" of the record.

TX 1255 cannot be admitted under Rule 803(6) because the trial record is devoid of satisfactory evidence that those documents "were kept in the course of a regularly conducted business activity and that it was the regular practice of the business to make such records." See Fagiola v. National Gypsum Co., 906 F.2d 53, 59 (2d Cir. 1990); Dell Pub. Co. v. Whe-don, 577 F.Supp. 1459, 1464 n. 5 (S.D.N.Y. 1984) (refusing to allow documents to be admitted under Rule 803(6) where foundation witness could not personally vouch that such documents were kept in the regular course of business).

Rule 803(6) provides that even if the hearsay statement meets the above requirements and thus constitutes a business record, the business record also may be excluded if the "source of information or the method or circumstances of preparation indicate lack of trustworthiness." Freidin, 849 F.2d at 720-21. As explained above, the excerpt of TX 1255 relied upon by Agfa does not appear to accurately reflect the disagreement between Kodak and DuPont at the 1995 meeting. This inaccuracy, Agfa's inability to identify the document's author or authors and the source of the hearsay statement, and questions pertaining to if and when the document may have been edited each indicate a lack of trustworthiness that mandates exclusion of TX 1255 as hearsay.

2. Deposition Testimony of Mr. Waaser

Federal Rule of Civil Procedure 32 (Fed.R.Civ.P. 32) and Fed.R.Evid. 804(b)(1) provide that in order for a non-party deposition to be utilized at trial for anything other than impeachment, there normally must be a showing of unavailability of the witness. Mr. Waaser was present at trial and testified. Nonetheless, Kodak designated portions of his deposition for inclusion in the trial record.

It is implicit from the Pretrial Order and pre-trial conference with counsel that designated testimony is automatically admitted into evidence, subject to objection by the other party. Agfa chose to designate testimony of Mr. Waaser. This designated testimony, together with any counter-designations made by Kodak under Rule 32(a)(4) are therefore in evidence, at least to the extent that Kodak's objections to Agfa's designations and Agfa's objections to Kodak's counter-designations are not sustained. Thus, by designating portions of Mr. Waaser's deposition, Agfa seemingly has waived objection to their admission.

Even if it has not waived its objection to the designations, there are no material facts in dispute that are solely and exclusively supported by portions of Mr. Waaser's deposition testimony. Furthermore, there was no fact material to the Report and Recommendation that was supported solely by a deposition designation of Mr. Waaser. Additionally, as noted below, allowing the admission of the designated deposition testimony expedited the trial. The deposition testimony served to supplement the trial testimony, trial testimony that permitted the credibility of the witness to be assessed.

3. Deposition Testimony of Dr. Hudson

Agfa did not designate any of Dr. Hudson's deposition testimony. To the extent that Kodak has designated Dr. Hudson's deposition testimony, Agfa has preserved its objection on the ground that Dr. Hudson was available at trial:

MR. HARTMANN: There is an objection to the deposition testimony of any witness that is available. The witnesses ought to be taken live when they're here, period, full stop.

Trial Tr. 462.

Federal Rule of Civil Procedure 32(a)(1) permits a party's use of a deposition "for any . . . purpose permitted by the Federal Rules of Evidence." This provision permits the substantive admission of depositions which fall within the ambit of Fed.R.Evid. 801(d)(2) irrespective of whether the deponent is available to testify at trial. See Globe Sav. Bank v. United States, 61 Fed.Cl. 91, 94-96 (2004); Long Island Sav. Bank v. United States, 63 Fed.Cl. 157, 163-64 (2004), rev'd on other grounds, 503 F.3d 1234, 2007 U.S.App. LEXIS 21915 (Fed. Cir. 2007).

Agfa attempts to distinguish Globe and Long Island on the grounds that in both cases the witnesses did not actually testify at trial, and the witnesses were employees at the time of the depositions. The first argument is not persuasive. Rule 801(d)(2) does not require a showing of the unavailability of the declarant. Globe, 61 Fed.C1. at 94. With regard to the second argument, both Globe and Long Island involved Rule 801(d)(2)(D), which applies where a statement is offered against a party and is "a statement by the party's agent or servant concerning a matter within the scope of the agency or employment, made during the existence of the relationship."^1a Kodak has not attempted to rely on 801(d)(2)(D), presumably because Dr. Hudson was not employed by Agfa at the time that the February 5, 2004 and October 12, 2004 depositions were taken, and Dr. Hudson was not receiving payments from Agfa as an expert prior to the depositions.

1a Rule 801(d)(2)(D) requires that the testimony (i) be made by the party against whom it is offered, (ii) concern a matter within the scope of the deponent's agency or employment, and (iii) be made during the existence of the deponent's relationship with the party. Globe, 61 Fed.Cl. at 97.

Rather, Kodak has relied on Rule 801(d)(2)(B) and (C), which make no mention of an "agent" or "servant". These rules apply where a statement is offered against a party and is:

(B) a statement of which the party has manifested an adoption or belief in its truth, or

(C) a statement by a person authorized by the party to make a statement concerning the subject, or

Kodak cites Glendale Fed. Bank, FSB v. United States, 39 Fed. Cl. 422, 425 (1997), for its holding that when a party puts forth an expert witness as a testifying expert at trial, the expert's prior deposition testimony in the same case constituted an admission against the party that retained him. Kodak argues that pursuant to this holding Agfa's presentation of Dr. Hudson as a testifying expert at trial means that his prior deposition testimony was authorized by Agfa under 801(d)(2)(C). However, in Glendale at the time that their testimony was taken these witnesses were "experts" and that was their only role in the litigation. Here, by contrast, Dr. Hudson was not engaged as an expert at the time of his deposition. His initial deposition was that of a fact witness.

Kodak argues that Agfa has manifested an adoption or belief in the deposition testimony under 801(d)(2)(B) and authorized the deposition testimony under 801(d)(2)(C) by virtue of (a) the sharing of the same legal counsel by Agfa and Dr. Hudson, and (b) Agfa's submission of Dr. Hudson's affidavit in support of Agfa's December 2003 summary judgment motion.

To the extent that Agfa has adopted or indicated a belief in statements made by Dr. Hudson in his depositions, such as in Agfa's December 2003 summary judgment motion, as provided under Fed.R.Evid. 801(d)(2)(B), these statements are admissible.

Kodak further argues that the Hudson depositions may be admitted under Fed.R.Civ.P. 32(a)(3)(E), which provides:

The deposition of a witness, whether or not a party, may be used by any party for any purpose if the court finds . . . upon application and notice, that such exceptional circumstances exist as to make it desirable, in the interest of justice and with due regard to the importance of presenting the testimony of witnesses orally in open court, to allow the deposition to be used.

Clearly Agfa had notice that deposition testimony was to be used, because the Pre-Trial Order expressly provided for designation, counter-designation and objections to designated deposition testimony. Thus, there was both application and notice. Moreover, in this instance, the deposition excerpts are not offered to the exclusion of live testimony but as a supplement to live testimony in order to shorten the trial time.

While due regard is to be given to live testimony, courts have looked to various factors in deciding whether to allow deposition testimony to be used as a substitute for live testimony by an otherwise available witness. In Borchardt v. U.S., 133 F.R.D. 547 (E.D.Wis. 1991), deposition testimony was permitted due to cost considerations. In SCM Corp. v. Xerox Corp., 76 F.R.D. 214, 216 n. 2 (D.Conn. 1977), the court expressly welcomed deposition testimony due to an already lengthy trial. Here, trial was scheduled to be concluded in seven (7) days, which of necessity was a very accelerated trial schedule considering the complex subject matter, number of patents, number of accused products, and number of issues to be litigated. While such a quick trial is the opposite of SCM, the underlying rationale remains the same; namely, a desire not to waste trial time. Certainly the interests of justice would be served by limited use of deposition testimony supplementing trial testimony in view of the accelerated schedule.

Moreover, the case involved infringement of three patents, including twenty claims thereof, and eleven accused products. Much of the testimony to be elicited would have been redundant, time-consuming and was not truly disputed evidence.

Furthermore, this case was not tried to a jury. The purpose of a trial is to resolve material disputes of fact, because otherwise summary judgment under Rule 56 would be available. Of the depositions to which Agfa objects, the witnesses testified for a period sufficient to assess their credibility. Furthermore, while Agfa objects to the admissibility of the Waaser and Hudson deposition excerpts, its post-trial submissions designated deposition excerpts from Dr. Booms and Messr. Van Dam, Dillenbeck and Hawley, all of whom testified at the trial. Agfa has not explained why it can designate the deposition testimony of witnesses who testified, but Kodak is precluded from doing so.

Also, almost without exception, the September 7, 2007 Report and Recommendation relied on multiple sources for the findings of fact. There was no fact critical to conclusions reached in the Report and Recommendation that relied solely on the fact deposition of either Mr. Waaser or Dr. Hudson.

CONCLUSION

Based on the foregoing, I recommend sustaining Kodak's objection to the admission of TX 1255. I furthermore recommend that the Agfa objections to the Waaser and Hudson fact depositions be overruled.

Sept. 27, 2007

REPORT AND RECOMMENDATION ON OBJECTED TO EXHIBITS

JOSEPH W. BERENATO, III, Special Master.

INTRODUCTION

Trial in this matter was held May 3-11, 2007. At the conclusion of trial, I requested the parties to attempt to reach agreement on the admissibility of the exhibits to which they had objected in the Pre-Trial Order. Thereafter, on June 14, 2007, the parties filed an Agreed Upon Exhibit list. Each party continued to object to exhibits propounded by the other. Although the parties were permitted an opportunity to expand upon the brief objections contained in the Pre-Trial Order, Kodak alone chose to elaborate on only certain of its objections to Agfa exhibits.

Having considered the objections made by the parties to the exhibits, the following represents my report and recommendation.

Agfa's Objections To Kodak Exhibits

Agfa's objections are overruled with the exception that its objections to Exhibits Nos. 288, 289, 380, 666, 870 and 927 are sustained.

Kodak's Objections To Agfa Exhibits

Kodak's objections to Agfa's exhibits are overruled except that the objections to Exhibits Nos. 1293 and 1386 are sustained. I previously have filed a Report and Recommendation concluding that Kodak's objection to Agfa's proposed Exhibit 1255 be sustained and nothing herein should be construed to the contrary.