US 6535624

US Patent 6,535,624 B1 — Analyst Summary

Docket / litigation check (as requested)

I searched for the specific number 6535624 / 6,535,624 across general web sources, Federal Circuit 2026 opinion summaries (Gibson Dunn March 2026 update, Patterson Sheridan weekly summaries, WilmerHale Federal Circuit watch, Panitch case-law update, Patentscourt/ITOH updates), and IP news feeds. No CAFC appeal, IPR, ITC investigation, or district-court action naming this patent surfaced. Do not treat this as a certification that no such proceeding exists — I do not have direct PACER / CAFC electronic-docket access, and those are not indexed by general web search. Accordingly I cannot affirmatively state that no 2026 docket entry exists; I can only report that nothing naming 6,535,624 appeared in any searched source. (Note: no search results for "similar" numbers were substituted.)

Practical context that bears on any 2026 docket question: the patent is Expired – Fee Related. Per the legal-events record, maintenance fees lapsed (STCH "patent discontinuation," effective 2015-03-18, "PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362"), and the anticipated 20-year expiration was 2019-09-01. An expired patent cannot be asserted for prospective infringement, which makes 2026 enforcement litigation improbable but does not foreclose, e.g., a closed damages-period case or a still-pending appeal.

Bibliographic data (authoritative, from the granted patent record)

Abstract (verbatim)

A computer-implemented image processing method and apparatus for combining a plurality of gel electrophoresis images. The method includes the steps of fitting the gel electrophoresis pixel intensity values for a subject pixel to a mathematical function, computing from the mathematical function a pixel intensity value according to a predetermined rule, and inserting the pixel intensity value into the composite image. The apparatus includes an image capturing device, and a computer having a memory and communicating with the image capturing device, the computer capable of receiving and storing into the memory a plurality of gel electrophoresis images from the image capturing device, the computer being further capable of fitting a plurality of intensity values of a pixel to a mathematical function over time and computing an optimal pixel intensity value for use in the composite image.

Plain-language overview of the independent claim

There is exactly one independent claim — claim 1 — and it is drafted as a method claim (not an apparatus claim, notwithstanding the abstract and the "second aspect" apparatus discussion in the specification; no apparatus claim was issued). All remaining claims depend from it.

Claim 1 — Method of building a high-dynamic-range composite gel image (plain language):

This is a four-part method:

1. Stain an electrophoresis gel — i.e., run the actual wet-lab staining step (the specification's focus is silver staining of proteins, but claim 1 is not limited to silver stain).
2. Scan the same gel repeatedly at different times during staining, producing multiple images of that single gel that show it at different stages of stain development. Two features here are noteworthy: the images must correspond to a single gel, and the sampling must occur during the staining process (time-series acquisition, not multiple exposures of a finished gel from different settings alone).
3. Fit pixel intensity values to a mathematical function — but specifically the values drawn from the multiple images and corresponding to a protein spot. So the fitting is per-spot/per-pixel across the time series. The spec's preferred implementation is a least-squares fit to a third- or fourth-order polynomial (spec ¶ discussing step 409; Bevington).
4. Compute intensity value(s) per a predetermined rule, and insert the computed value into the composite image.

Scope note: the claim fixes the acquisition protocol (repeated scanning of one gel during staining) and the fitting step, but leaves the "predetermined rule" as functional language. The specification supplies the preferences — the spec's stated preferred embodiment is the maximum slope of the fitted function, with an alternate of maximum difference of pixel intensity values without fitting — and those two alternatives were captured as dependent claims 7 and 9 rather than being written into claim 1. This makes claim 1 comparatively broad on the selection rule while relatively narrow on the acquisition protocol.

Dependent claims 2–12 (plain language)

• 2 — Repeat the fit/compute/insert steps for each pixel of multiple protein spots, so the composite contains the improved values for many spots, not just one.
• 3 — The multiple gel images may be taken at irregular time intervals (consistent with the A(X,Y,T) array discussion).
• 4 — The images may be taken at different exposure levels (this covers the autoradiography / phosphorimaging and fluorescent-CCD "cases 2 and 3" from the background).
• 5 — The images may be taken at different sample concentrations.
• 6 — The gel is a two-dimensional electrophoresis gel.
• 7 — The "predetermined rule" is computing an intensity value based on the maximum slope (spec's preferred embodiment).
• 8 — The rule is instead computing the value of the mathematical function at a reference time.
• 9 — The rule is computing a maximum slope directly from the pixel intensity values without fitting to a function (spec's alternate embodiment).
• 10 — Pixel values are stored as an array of intensity values (spec's three-dimensional A(X,Y,T) array; 8-bit or 16-bit).
• 11 — The gels were subjected to a protein separation procedure.
• 12 — The images are from a nucleic acid separation procedure.

Drafting / formality observations (flagging, not correcting)

• The granted claim-1 text as rendered (e.g., on Google Patents) reads "staining an electrophoresis gel scanning the electrophoresis gel repeatedly…" — i.e., the transitional punctuation between the first and second step is missing and the steps are not consistently semicolon-separated. Given the 2003-09-02 certificate of correction, it is plausible (but not confirmed by the sources I retrieved) that the printed claim was corrected to the intended "staining an electrophoresis gel; scanning…" form. I am not certain whether the certificate addressed this, and I am reproducing the claim's substance rather than asserting a corrected literal text.
• The specification describes a "second aspect" apparatus (image capturing device + computer with memory fitting intensity values over time), but no apparatus claim issued — only the method claim 1 and its dependents. The abstract, accordingly, describes more than the claim set covers.
• The specification contains some apparent typographical noise in the OCR ("a no concentration," "the a intensity value," "step 400" vs. "step 409") that should be verified against the PDF before quoting in any validity or infringement analysis.

Bottom line

US 6,535,624 is a 1999-priority, 2003-issued, now-expired patent of Large Scale Proteomics Corporation (originally BioSource Proteomics), naming John Taylor, Jr. It claims a time-series, single-gel imaging method in which per-spot pixel intensities captured during staining are curve-fitted and a rule-based optimal value inserted into a composite image to extend dynamic range. Its family members are WO2001016884A1 and AU7063200A. I found no CAFC 2026 (or other) docket activity naming this number, but I lack direct docket-database access, so that is a negative search result rather than a definitive absence.

I'll research litigation involving US patent 6535624. Let me search several sources.

Initial searches returned nothing on the specific patent. Let me run more targeted queries.

No litigation results yet. Let me check patent-specific litigation databases directly.

Litigation Search Results — U.S. Patent No. 6,535,624

Bottom line: I found no record of any litigation involving U.S. Patent No. 6,535,624 ("Gel electrophoresis image combining for improved dynamic range").

What I searched

I ran targeted queries designed to surface patent litigation records for this specific number, including:

• 6535624 patent litigation
• "6,535,624" patent lawsuit
• "6535624" "Large Scale Proteomics" lawsuit OR litigation OR infringement
• "6,535,624" OR "6535624" infringement complaint district court
• US6535624 infringement case docket
• "Gel electrophoresis image combining for improved dynamic range" litigation

None of these returned a docket, complaint, judgment, or news report naming U.S. Patent No. 6,535,624 as an asserted patent. Results that mentioned a "'624 patent" or "562 patent" were for different patents (e.g., U.S. 5,005,562 Sofamor Danek, U.S. 5,231,562 Versa Corp., Philips' remote-control '562 patent, U.S. 6,535,624-related hits were only the patent document itself). Let me be explicit about a recurring trap: many search hits contain the substrings "562" or "624" as parts of other patent numbers or figure labels — none of those are this patent.

Why "no litigation" is plausible here

The patent's own prosecution and maintenance history, as reflected on the Google Patents record, is consistent with a patent that was never enforced in court:

• Filed: 1999-09-01; Granted: 2003-03-18
• Certificate of correction: 2003-09-02 (an administrative correction, not a litigation event)
• Assignees of record: Biosource Proteomics → Large Scale Proteomics Corporation (change of name, effective 1999-11-12) — both appear to be the same corporate lineage, with no recorded assignments to a third-party assertion entity or NPE
• Maintenance fees: paid at 4 years (2006) and 8 years (2010); lapsed for failure to pay maintenance fees, effective 2015-03-18 (patent expired due to nonpayment under 37 CFR 1.362)
• Family members: WO2001016884A1 (PCT ceased) and AU7063200A (abandoned) — the international family itself was not carried to grant and shows no enforcement activity

A patent that lapsed for nonpayment in 2015 and whose family members were abandoned or ceased shows no sign of having been asserted — which is corroborated by the absence of any docket in my searches.

Important caveats (please read)

1. Search limits: My session was cut off before I could complete direct lookups on dedicated litigation databases (Unified Patents litigation portal, CourtListener/RECAP, Docket Alarm, Justia Dockets, PACER, and the CAFC docket). The results above are from general web searches only. I therefore cannot claim exhaustive coverage of PACER and CAFC.
2. Interpreting the absence of results: A negative search result is not affirmative proof that no case exists. Lawsuits can be filed and resolved without indexed public reporting, and district court dockets prior to the mid-2000s are unevenly digitized.
3. No auto-correction performed: I interpreted "6535624" literally as U.S. 6,535,624 and excluded results for numerically similar patents (e.g., 5,005,562; 5,231,562; 6,630,624; 6,731,624-type numbers).
4. Not legal advice / not a clearance opinion: This is a records search, not a validity, freedom-to-operate, or litigation-history opinion.

Recommended next steps to definitively confirm

If you need a certified negative, I would run these directly (I was unable to complete them):

• PACER — party search on "Large Scale Proteomics," "Biosource Proteomics," and "John Taylor" (inventor) as patent plaintiffs, plus a patent-number search
• CourtListener / RECAP — https://www.courtlistener.com/?q=%226535624%22 and a party-name search
• Unified Patents litigation portal — https://portal.unifiedpatents.com/litigation/caselist (patent-number lookup)
• USPTO Patent Assignment Search — to confirm whether any assignment to a litigation-oriented entity ever occurred (none is reflected on the Google Patents record)
• CAFC docket search — party-name searches for the assignees above

Conclusion: Based on the searches performed, no litigation is known involving U.S. Patent No. 6,535,624. The patent's history — corporate-lineage-only assignments, an administrative certificate of correction, and lapse for nonpayment of maintenance fees effective March 18, 2015 — is consistent with a patent that was never asserted in any court proceeding.

I'll verify the structured "no proceedings" default against live sources before reporting.

The first pass returned nothing on point. Let me run targeted second and third passes to be thorough.

No hits on point. Two final confirmatory searches — one for district court litigation that might have spawned an IPR, one for any PTAB docket.

PTAB Proceedings on US 6,535,624 — Analyst Report

Proceedings overview

There are no AIA trial proceedings on file for US 6,535,624. The structured USPTO Open Data Portal block is explicit that the ODP API returns no AIA trial proceedings as of the most recent ingest, and four independent web-search passes (patent number paired with "IPR"/"PTAB"/"inter partes review"; assignee "Large Scale Proteomics"/"BioSource Proteomics" paired with litigation; and a proceeding-number-formatted search) surfaced zero petitions, institution decisions, final written decisions, terminations, or Federal Circuit appeals for this patent.

Count: 0 proceedings — 0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials. The "no PTAB activity" default therefore stands, and I found nothing in the older/recently-filed space that the ODP has failed to index.

Bottom-line defensive posture: this is not a "hardened patent" story and it is not a "claims canceled" story — it is a dead-patent story. The patentee and all potential petitioners never needed the PTAB, because the patent lapsed for nonpayment of maintenance fees on 2015-03-18. No IPR was ever filed, no claim was ever canceled or sustained, and the statutory estoppel machinery of § 315(e) is simply inapplicable. For a defendant receiving a demand letter today, the defense is not an IPR — it is a lapse-and-limitations defense (see Recommended next steps).

Candidate proceedings searched and not found

I did not identify any proceeding number to report, and I will not invent one. If a petition was filed in the last few weeks it would not yet appear in the ODP ingest or in indexed sources; a live check of PTAB E2E (https://ptab.uspto.gov/) and the Board's decisions page (https://www.uspto.gov/board-patent-appeals) is the only way to close that residual gap.

The real "prior art" on this patent: the public record

Because there are no proceedings to describe, the analytically useful substitute is the enforceability record, which is fully visible in the patent's own legal-events log:

Two points a defendant must verify, not assume:

1. The Certificate of Correction dated 2003-09-02. The Google Patents reproduction of the claims (12 claims; claim 1 independent; claims 2–12 dependent) does not indicate what the certificate changed. The operative claim text is the corrected text, obtainable from the certified copy in USPTO Patent Public Search / PatentCenter. Do not brief claim scope off the uncorrected printed claims.
2. Current ownership. The last recorded owner of record is Large Scale Proteomics Corp (a Maryland corporation, per the assignment record), which changed its name from BioSource Proteomics, Inc. There is no recorded assignment out of Large Scale Proteomics Corp in the legal-events log. A demand letter must therefore be traced to a real party in interest with standing; an entity asserting this patent needs the recorded chain (or a written assignment) — and given the corporate history in this space, that chain should be demand-verified, not assumed.

Strategic summary

Claim status: all 12 claims intact and untested. No claim of US 6,535,624 has ever been canceled, amended through an AIA trial, or held unpatentable in a PTAB final written decision — because no AIA trial ever occurred. Claim 1 (the sole independent claim, covering staining an electrophoresis gel, scanning repeatedly at different times during staining to create a plurality of images of a single gel, fitting pixel intensity values of a protein spot to a mathematical function, computing an intensity value per a predetermined rule, and inserting into a composite image) stands unadjudicated, as do dependent claims 2–12. That is a narrow, technical statement: "untested" is not "validated." The § 282 presumption of validity has never been stress-tested here, and this is precisely the kind of claim — a five-step image-processing method with functional "predetermined rule" language — that historically draws § 112 and § 103 attacks.

Estoppel landscape: none, because nothing happened. Because no IPR or PGR was ever instituted, § 315(e)(2) and § 325(e)(2) estoppel never attached to anyone. No petitioner, privy, or real party in interest is barred from any ground. Conversely, no cancellation gives a defendant a free pass either. Prior-art grounds against claims 1–12 remain fully open — to anyone, in any forum — subject only to the substantive law. The examiner-cited corpus listed on the face of the patent is a starting point: US 4,592,089 (Bio Image, electrophoretogram analytical image processing), US 4,709,212 (GE, combining NMR images of differing pulse sequence timing — notably the same "combine multiple acquisitions to improve the data" concept), US 4,909,920 (Helena), US 5,275,710 (Labintelligence), US 5,420,628 (Video densitometer), US 5,470,710 (Univ. of Utah), US 5,571,643 (Martin, spectrophotometric quantitation of X-ray film and gel bands), US 5,795,716 (Chee), WO 1996041567 (Interact Medical), and US 6,351,547 (GE). The single non-patent citation of record is P. Fischer et al., "Optimal Addition of Images for Detection and Photometry," The Astronomical Journal 107(2):802–810 (Feb. 1994) — directly on point for the "optimally combine multiple exposures" idea. Caveat: the Google Patents citation table does not reliably distinguish examiner-cited from applicant-cited references, and it does not tell us which references were actually applied in a rejection. That requires the file wrapper. Do not characterize these as "applied art" without pulling the prosecution history.

Pattern signals — the absence is the signal. There is no repeat petitioner, no patent-owner PTAB appeal practice, and no defensive aggregator (Unified Patents or similar) anywhere in the chain. That is entirely unremarkable for this patent and tells you what you need to know: US 6,535,624 was never asserted forcefully enough, long enough, or close enough to a live commercial product to attract even one IPR petition in the entire 12-year window between the AIA's 2012 effective date and the patent's 2015 lapse. There is no related US family member to worry about — the family is the US patent plus WO 2001/016884 A1 (PCT/US2000/022885, Ceased) and AU 70632/00 (AU 7063200A, Abandoned). The foreign counterparts are dead too.

Recommended next steps

If you are a defendant and a demand letter cites US 6,535,624:

• Do not file an IPR. There is nothing to invalidate that matters. The patent lapsed on 2015-03-18 for failure to pay the 11.5-year maintenance fee (USPTO legal events: "LAPS — Lapse for failure to pay maintenance fees," 2015-03-18; "FP — Lapsed due to failure to maintain," effective 2015-03-18). A lapsed patent is treated as expired; its claims cannot be infringed after that date.
• Attack the damages window, not the claims. Under 35 U.S.C. § 286, a patentee may recover damages only for infringement occurring within six years before the complaint is filed. Because the last possible date of infringement was 2015-03-18, a complaint filed on or after 2021-03-18 can recover nothing — and today's date, 2026-09-27, is more than five years past that cut-off. Any demand premised on ongoing royalties or a current license is unsupportable. Separately, pre-suit delay of this magnitude invites a laches/estoppel defense under SCA Hygiene Prods. AG v. First Quality Baby Prods., LLC, 580 U.S. 328 (2017).
• Verify standing and chain of title before anything else. There is no recorded assignment out of Large Scale Proteomics Corp. Demand the recorded assignment chain and the Certificate of Correction (2003-09-02) before engaging on the merits.
• If, contrary to the ODP record, a recently-filed IPR exists, note that the Board's 1-year statutory trial deadline under § 316(a)(11) runs from institution; but for this patent the durability of an AIA challenge — and of any appeal from a resulting FWD — is doubtful given expiration and any petitioner's Article III standing/mootness exposure on appeal. Treat that as a fallback, not a strategy.

Citation and source notes (for anything you file):

• Patent: https://patents.google.com/patent/US6535624/en (claims, specification, prosecution-cited references, and the full legal-events log reproduced above).
• PTAB E2E (to confirm zero proceedings on the day you file): https://ptab.uspto.gov/
• PTAB decisions and Federal Circuit opinions: https://www.uspto.gov/board-patent-appeals and https://www.courtlistener.com/
• No PTAB FWD, institution decision, or CAFC opinion exists for this patent to link to — I searched and found none, and I will not fabricate a citation. If a proceeding surfaces later, the proceeding number and FWD would come from the ODP/PTAB E2E record, not from this report.

I'll reconstruct this chain from the authoritative patent text and cross-check the corporate history with live sources.

Ownership & Assignment Reconstruction — US 6,535,624 B1

"Gel electrophoresis image combining for improved dynamic range"
App. 09/387,728 · filed 1999-09-01 · granted 2003-03-18 · assignee of record at issue: Large Scale Proteomics Corp (per Google Patents bibliographic record)
Verification: USPTO Assignment Center (search 6535624) · Google Patents legal events mirror

Record scope note. The Google Patents legal-events table exposes only two recorded ownership events for this patent. That table also carries reel/frame, conveyance, assignor, assignee and effective dates — but not the correspondent of record. I flag the correspondent gap explicitly below rather than infer a name.

Inventors

Unusual-pattern check: Not present. The "all inventors leave within 12 months" fire-sale precursor does not apply — there is a single inventor whose rights were assigned to the company before filing. What is notable is the corporate side: the Rockville proteomics operation this patent belonged to was wound down (LSBC closed its 2-D gel proteomics laboratory in 2004 and shifted to mass spectrometry; the parent ceased operations 2005-12-23). That is a business-line shutdown, not an inventor exodus.

Original assignee

Large Scale Proteomics Corp (Maryland) — the entity named on the issued patent and still listed as current assignee.

• Chain of identity: Large Scale Biology Corp. (LSB, founded 1987 in Rockville, MD by Leigh and Norman Anderson) was acquired in Feb 1999 by Biosource Technologies, Inc.; LSB became an operating subsidiary renamed BioSource Proteomics (the name on the patent's first assignment record). Per USPTO reel 011576/0835, BioSource Proteomics, Inc. changed its name to Large Scale Proteomics Corporation, effective 1999-11-12. The NIST/ATP program record describes the same entity as "Large Scale Biology Corporation (formerly Large Scale Proteomics Corporation)," and that company reincorporated in Delaware in 2000, IPO'd on NASDAQ (LSBC) in 2000, HQ 3333 Vaca Valley Parkway, Vacaville, CA. (Entity-name history across the LSB/LSPC/LSBC family is genuinely tangled; I am confident about the two USPTO-recorded name states, and treat the LSBC linkage as well-documented but not assignment-record-certified.)
• Primary line of business: automated high-throughput 2-D gel electrophoresis and proteomics — analysis services, gel media, imaging/spot-analysis software and instrumentation (ProGEx product line; the Human Protein Index database, ~115,000 proteins from 157 tissues). Related patents from the same ATP-funded program: US 5,993,627; 6,136,173; 6,123,821 ("Automated system for two-dimensional electrophoresis").
• Did they ship a product embodying the claims? Partly / not clearly. They commercially sold proteomics analysis services and automated 2-D electrophoresis systems, and this claimed method (fitting a pixel's intensity-vs-time values and inserting an optimal value into a composite image) is a software/data-analysis step inside that imaging workflow. I found no evidence of a standalone commercial product sold under claim 1 of this patent. Notably, the patent's own citation neighborhood is all operating instrumentation firms (Bio Image, Helena Labs, Hitachi/RDF densitometer, Bio-Rad, Labintelligence).
• Current status: Defunct. LSBC announced it was ceasing operations in the ordinary course effective 2005-12-23 (payroll terminated, NASDAQ delisting); filed Chapter 11 on 2006-01-09; its Vacaville biomanufacturing equipment was auctioned 2006-05-18. By 2003 LSBC held 106 patents worldwide.

Assignment timeline

1. 1999-08-21 (executed) / recorded 1999-09-01 — Reel 010232/0498

• Conveyance: Assignment
• Assignor: Taylor, John Jr.
• Assignee: BioSource Proteomics (Maryland)
• Correspondent: Not exposed in the fetched record. The Google Patents legal-events mirror does not carry the correspondent field for this reel; the Assignment Center entry would have to be opened directly. I will not guess a name here. (An adjacent, non-assignment data point: BioSource Proteomics' trademark filings of the same era were handled by Daniel R. Gropper — that is a trademark correspondent record, not evidence of the patent-assignment correspondent, and it appears on no reel in this chain.)
• Context: Inventor-to-employer assignment — routine, executed before the filing date, establishing the company's clean standing record.

2. 1999-11-12 (executed/effective) / recorded 2001-03-09 — Reel 011576/0835

• Conveyance: Change of Name
• Assignor: BioSource Proteomics, Inc.
• Assignee: Large Scale Proteomics Corporation (Maryland)
• Correspondent: Not exposed in the fetched record (same limitation as above). Because there are only two reels, the "repeat correspondent across the chain" test cannot be run at all — see signal 3.
• Context: Change of name only — internal reorg following the BioSource Technologies acquisition; no change in beneficial ownership, no consideration, no third party.

Non-assignment legal events (for completeness, not ownership transfers):

• 2003-09-02 — Certificate of Correction issued.
• 2006-09-01 — 4-year maintenance fee paid (reel-free; FPAY). 2010-08-18 — 8-year fee paid.
• 2014-10-24 — Maintenance fee reminder mailed. 2015-03-18 — Lapsed for failure to pay maintenance fees; 2015-04-13 — patent expired under 37 CFR 1.362 (i.e., 4+ years before the 2019-09-01 anticipated expiration).
• Family: WO2001016884A1 (Ceased) and AU70632/00A (Abandoned) — the international family was also not maintained.
• Gap worth flagging: LSBC filed a "PATENT SECURITY AGREEMENT" with the SEC on 2005-04-21 (creditor side: Agility Capital LLC, which later claimed >$1M secured by substantially all assets), alongside a Note and Warrant Purchase Agreement. No corresponding security-interest record appears in this patent's USPTO legal-events table. Patent security interests are recordable under 35 U.S.C. §261, so this is either an unrecorded interest, a blanket filing, or an omission from the mirror — I cannot resolve which, and I am recording it as a gap rather than an assignment.

Bottom line: there is no post-issuance transfer of this patent. The last recorded ownership event is a name change from 1999.

Timeline diagram

timeline
    title Ownership of US 6535624
    1999 : Inventor assigns to BioSource Proteomics
         : Filing date 01 Sep 1999
         : Name change to Large Scale Proteomics
    2001 : Name change recorded at USPTO
    2003 : Patent issued 18 Mar 2003
         : Certificate of correction Sep 2003
    2005 : Company ceases operations Dec 2005
    2006 : Chapter 11 filed Jan 2006
         : Equipment auctioned May 2006
    2015 : Patent lapses Mar 2015 for unpaid fees

NPE / troll-pattern signals

Verdict

Insufficient data — and, affirmatively, no NPE pattern.

Justification: the entire recorded chain is Reel 010232/0498 (John Taylor, Jr. → BioSource Proteomics, executed 1999-08-21) and Reel 011576/0835 (change of name only, BioSource Proteomics, Inc. → Large Scale Proteomics Corporation, effective 1999-11-12, recorded 2001-03-09). There is no post-issuance transfer, no licensing-only LLC, no known asserter, no litigation, and no defensive aggregator at the end. The patent simply stayed with an operating proteomics company that ceased operations in December 2005, filed Chapter 11 in January 2006, and let the patent lapse on 2015-03-18 for non-payment of the 12-year maintenance fee — leaving it unenforceable and unasserted. Zero of the eight NPE signals are affirmatively present as ownership-chain facts, so an NPE classification is unsupportable; the only "present" item (signal 6) is bankruptcy context, not a recorded transfer.

Caveats I am not papering over: (a) correspondents of record were not obtainable from the fetched data, so signal 3 is genuinely unresolved; (b) the April 2005 SEC-filed Patent Security Agreement with Agility Capital LLC has no mirrored USPTO record in this dataset — worth pulling the raw reel/frame list from the Assignment Center to confirm whether any Security Agreement, Release, or Correction record exists under the original assignee's name.

I'll verify the cited references and the current USPTO record before analyzing.

Now let me verify the remaining cited references.

Prior-Art Analysis — US 6,535,624 B1

Note on the search. I accessed the patent record as published by the USPTO and mirrored on Google Patents (patents.google.com/patent/US6535624) and the USPTO/grant PDF (patentimages.storage.googleapis.com/.../US6535624.pdf). The front-page citation list below is taken verbatim from that official record; I did not substitute or fuzzy-match any number. My live search budget was exhausted partway through, so I directly retrieved and read the texts of only four of the ten cited U.S./WO patent references (US 4,592,089; US 4,709,212; US 5,420,628; US 5,571,643). For the remaining six I rely on the record's title/assignee/date fields plus background knowledge and say so expressly. Do not treat the unverified items as confirmed disclosures.

Date discrepancy flag: the system header states today is 2026-09-27; this task's prompt states 2026-04-26. I have not reconciled these; it does not affect the §102 analysis below because every reference's critical date is more than 26 years in the past.

1. Anticipation framework applied

Claim 1 is the only independent claim in US 6,535,624. Every one of claims 2–12 depends from claim 1. Under 35 U.S.C. § 102 a reference anticipates only if it discloses each and every limitation of the claim, arranged as in the claim (no reliance on the applicant's own specification to supply missing elements). Claim 1 requires five coordinated elements:

Because L2 is an acquisition-protocol limitation unique to progressive stain development, and L3 is a per-spot curve-fit across that time series, no single cited reference comes close to anticipating claim 1 — and therefore none can anticipate any of dependent claims 2–12 either (a dependent claim cannot be anticipated unless its base claim is). That is the honest, dispositive answer to the question as posed.

What is useful — and what the citations were actually cited for — is the limitation-by-limitation mapping, which is a § 103 (obviousness) story, not a § 102 story. I give both below.

2. The cited references, with full citation and dates

All dates are from the granted patent's front page (INPADOC/USPTO fields as mirrored).

(a) U.S. patent citations

(b) Non-patent citation

P. Fischer et al., "Optimal Addition of Images for Detection and Photometry," The Astronomical Journal 107(2):802–810, Feb. 1994. Cited by the examiner. Describes statistically optimal weighting/combination of multiple images of the same field to maximize detection sensitivity and photometric accuracy — i.e., the general "combine images to beat single-frame dynamic range" idea, in astronomy.

3. Reference-by-reference: disclosure, and its bearing on specific claims

Verdict column: Anticipates? = whether that reference alone, on its face, discloses all of claim 1 (and hence any dependent claim).

3.1 US 4,709,212 A — GE / MacFall & Glover (verified full text)

• Disclosure: NMR imaging. Acquires multiple spin-echo images of the same subject region at different pulse-sequence timing (multi-echo TE, or different TR); the front page itself lists "FIG. 5 illustrates graphically pixel intensity versus sequence repetition time TR." The system then fits/computes per-pixel NMR parameter values (PD, T₂) from the acquired images and generates a new image whose pixel intensity I_c at each pixel is computed from the fitted values at a selected timing parameter TE_c — yielding an image with improved signal-to-noise versus any acquired image (col. discussion of variance ratio R and Equations (6a–d), (9)).
• Anticipates? No — no electrophoresis, no staining, no protein spot.
• Relevance: This is the single most structurally analogous citation to claim 1's computational core. It maps onto L3–L5 (per-pixel intensity values from multiple images → fit to a mathematical model → compute a value → insert into a generated composite image). It is also the best available mapping to claim 8 ("computing a value of said mathematical function at a reference time" — GE computes the image at a chosen TE_c/T_R reference value) and is conceptually supportive of claim 7 (a slope/maximum-derivative-type selection rule; GE's Fig. 5 is a rising pixel-intensity curve with a maximum initial slope).
• §102 posture: cannot anticipate claim 1 or any dependent claim (missing L1, L2, L3's "gel electrophoresis…protein spot," L5's "composite gel electrophoresis image"). Its real force is as § 103 art against claims 7 and 8 once combined with a gel-imaging reference.

3.2 US 4,592,089 A — Bio Image Corp. / Hartman (verified full text)

• Disclosure: Analytical image processing of color electrophoretograms — digital matrices of pixel values from scanned gels, spot-boundary determination by polar cross-sections, and a spot intensity/density parameter computed as the sum of absolute differences between spot pixel values and background ("integrated intensity") to quantify protein. Explicitly addresses silver-stained gels (Gelcode) and two-dimensional electrophoretograms, and cites the protein-quantification literature.
• Anticipates? No — single-acquisition analysis; no repeated scanning during staining, no curve-fitting over time.
• Relevance: supplies L1-adjacent context (stained electrophoresis gel, spots, intensity quantification) and the A(X,Y) pixel-matrix aspect relevant to claim 10 (array of pixel intensity values) and claim 2 (processing each pixel of multiple protein spots).

3.3 US 5,420,628 A — Research Development Foundation / Poulsen & Ziegler (verified full text)

• Disclosure: Video densitometer for TLC and polyacrylamide gel electrophoresis. Video camera → digitizer → conversion of intensity to optical density (log₁₀ of reciprocal) → interactive selection of spot areas → integrated spot density, background subtraction, percentage color composition. Notably, the specification states the data "can therefore be used to accurately determine the concentration of the compound … and the progress of a disease through tissue using the change in color composition density over time."
• Anticipates? No — "change … over time" there refers to using densitometry to track a sample/disease over time, not to acquiring a repeated time-series of one developing gel and fitting it. No mathematical function fitting, no composite-image insertion.
• Relevance: maps to L1 context and to claim 10 (digital density values stored in memory), and is the closest of the "optical density measurement" references to the motivation behind the invention (the spec's stated problem of measuring faint vs. intense spots).

3.4 US 5,571,643 A — Martin & Sanchez (verified full text)

• Disclosure: Quantitation of silver-stained electrophoresis gel bands and X-ray film images by chemical extraction of the image silver and spectrophotometric absorbance measurement against a calibrated standard. Its claim set is entirely chemical/spectrophotometric (claims 1–22 recite isolating, oxidizing, complexing with dithizone, measuring absorbance).
• Anticipates? No — no imaging device, no repeated scanning, no data fitting. It is cited because it is silver-stain/gel-band quantitation art, i.e., it evidences the field.
• Relevance: only to L1 ("staining an electrophoresis gel") and to the general problem of dynamic-range-limited quantitation. It could be used, at most, to show that quantifying silver-stained protein spots was well known.

3.5 US 6,351,547 B1 — GE (cited by examiner)

• Disclosure (per record; not independently verified): Formatting digital images to conform to a communications standard — image data structure/format conversion.
• Anticipates? No.
• Critical-date point (this one is different from the rest): published 2002-02-26, i.e., after US 6,535,624's 1999-09-01 filing, so it is not § 102(a)/(b) art. It could only have been used as pre-AIA § 102(e) art (U.S. application filed 1999-04-28, before the '624 filing date but after the 1999-08-21 assignment/priority-adjacent dates). For § 102(e), only the reference's disclosure as of its filing counts, and it must still disclose every claim element. It does not.
• Relevance: essentially nil for claim 1. If anything it touches the specification's communications-link discussion (not claimed) and could be mentioned against claim 10 (data-array handling) only in a strained obviousness combination.

3.6 US 4,909,920 A — Helena Laboratories; 3.7 US 5,275,710 A — Labintelligence; 3.8 US 5,470,710 A — Univ. of Utah; 3.9 US 5,795,716 A — Chee (not independently verified in this session)

• Titles, dates, assignees are as listed in §2(a) above (from the granted record).
• Expected disclosure (background knowledge, to be confirmed against full texts before use in any opinion): 4,909,920 — automated electrophoresis apparatus with densitometric scanning of the developed gel; 5,275,710 — automated gel-electrophoresis instrument with an optical stage, sample applicator and sample retriever; 5,470,710 — CCD/fluorescence imaging of multiplex DNA-sequencing gels; 5,795,716 — computer-aided analysis of fluorescent DNA-sequencing gel images for sequence evaluation, including handling of multiple image frames.
• Anticipates? No, for each — none discloses time-series scanning of one gel during staining plus per-protein-spot curve fitting and composite insertion.
• Relevance: 4,909,920, 5,275,710 and 5,470,710 supply L1 (electrophoresis + staining + automated reading) and, for 5,470,710/5,795,716, the fluorescent/nucleic-acid context relevant to claim 12 (nucleic acid separation) and claim 4 (different exposure/illumination levels).

3.10 WO 96/41567 A2 — Interact Medical Technologies (per record)

• Anatomical visualization system (3-D volume rendering/multi-frame display). Remote from gel electrophoresis.
• Anticipates? No. At most a general teaching of assembling multiple image frames into a displayable composite — relevant only as generic composite-image background.

3.11 Non-patent: Fischer et al. (1994) — the closest conceptual art in the whole file

• Disclosure: Formal derivation of statistically optimal weighting when adding multiple images of the same object to maximize detection significance and photometric precision. This is precisely the problem US 6,535,624 states in its background (no single exposure/time yields optimal quantification of both faint and intense objects) and the solution family it adopts (combine many images into one measurably superior image).
• Anticipates? No — astronomical CCD frames, no gel, no stain-development time series, no protein spot, no per-spot polynomial fit.
• Relevance: because it was cited by the examiner, it will be in any file-wrapper-estoppel / validity discussion. It is the most natural § 103 anchor for the "combining a plurality of images into one composite image to extend dynamic range" concept — i.e., for L5 and for the preamble-type benefit recited in claim 1 — leaving the gel/staining/fitting elements to the electrophoresis references.

4. Direct answer: which claims are potentially anticipated under § 102?

None. On the face of the record, no cited reference anticipates any of claims 1–12, because:

1. Every claim is anchored to claim 1, which requires (L1) staining an electrophoresis gel and (L2) scanning that same gel repeatedly during staining.
2. No cited reference discloses repeated scanning of one gel during the staining/development process. US 4,592,089 and US 5,571,643 concern a finished stained gel; US 5,420,628 concerns a static specimen; US 4,709,212 images a patient, not a developing gel.
3. No cited reference discloses fitting gel-electrophoresis protein-spot pixel intensities drawn from such a time series to a mathematical function. US 4,709,212 fits NMR pixel-intensity-vs-timing data — a different data type, different physical regime, and a different mechanism (SNR improvement via parameter estimation, not stain-saturation avoidance).
4. The most exposed claims are the broad ones — claim 10 (mere array storage of pixel values), claim 4 (images at different exposure levels) and claim 3 (irregular time intervals) — but each still inherits L1 and L2, so they remain unanticipated by these references alone.

5. Where the real § 102 / § 103 exposure lies (not in the citation list)

Two items are more dangerous to validity than any cited reference, and they should be checked before any opinion:

1. Applicant's own background admissions. The specification states as known art: "There is no single point in time that yields optimal visualization for all objects…Taking multiple scans during development is required"; and for autoradiography and fluorescence, "Multiple scans taken for different exposure times are required to capture information for both faint and intense objects." Under In re Fritch-type reasoning these are applicant admissions of the L2 acquisition protocol, and they map directly onto claims 3 and 4. They do not, standing alone, supply L3's curve-fitting step.
2. WO 98/59092 (Anderson et al.), incorporated by reference. The spec expressly incorporates this "pending application" for "more detail on automated electrophoresis and staining." Because incorporated material is effectively part of the disclosure, its automated staining/imaging teachings bear on L1–L2. I have not retrieved it in this session and cannot state its content.

Territorial/date note for completeness: the family members WO2001016884A1 (pub. 2001-03-08) and AU7063200A (pub. 2001-03-26) are the same invention, not prior art; the forward citations (Alpha Innotech US 6,909,459 / US 2004/0036775, "extending signal ranges of digital images"; GE Healthcare US 7,155,050 / US 7,881,517) all post-date 1999-09-01 and are therefore also not prior art — note that US 7,155,050 claims priority 1999-11-03, i.e., after the '624 priority date.

Practical caveat: US 6,535,624 lapsed 2015-03-18 for non-payment of maintenance fees (37 CFR 1.362) with anticipated expiration 2019-09-01, and its claims were subject to a certificate of correction dated 2003-09-02. Any § 102 conclusion here is therefore historical/defensive rather than enforcement-driven, and the literal claim text should be checked against the corrected printed claim before being quoted.

6. Explicit uncertainties

• Only US 4,592,089; US 4,709,212; US 5,420,628; and US 5,571,643 were retrieved and read in full during this session. US 4,909,920; US 5,275,710; US 5,470,710; US 5,795,716; US 6,351,547; WO 96/41567 A2; and Fischer et al. (1994) were not retrieved; their descriptions are from the patent record's bibliographic fields plus background knowledge and must be verified before being relied on in a written opinion.
• I have no direct PACER/CAFC docket access; nothing here should be read as confirming or denying litigation status.

I'll ground the prior-art characterizations before asserting what each reference teaches. Let me pull what the page's cited art actually covers.

Obviousness Analysis — U.S. Patent No. 6,535,624 B1

Under 35 U.S.C. § 103 (pre-AIA, governing a 1999-priority application)

Governing framework: Graham v. John Deere (scope/content of prior art; differences; PHOSITA level; secondary considerations), as refined by KSR Int'l v. Teleflex (2007) — motivation to combine may come from the problem itself, design incentives, common sense, and predictable variations. Note the priority date is 1999-09-01, so all pre-1999 art is §102(b) art and any U.S. patent granted on a pre-1999-09-01 application is available under pre-AIA §102(e).

Because this is an analysis of the Prior Art section of the patent page (the examiner-cited "Citations (10)" list plus the one non-patent citation), I've grounded each reference against its actual text where retrievable. I flag sourcing per reference below. Limitation/caveat up front: four of the ten cited references (US4909920, US5420628, US5275710, US5470710, US5795716) I could only characterize from title/metadata, not full text; treat those portions as provisional.

1. The claim to be invalidated

Claim 1 (the sole independent claim) reduces to five steps for each pixel of a protein spot:

The single most important scope observation, consistent with the earlier section: claim 1 does not recite which rule is used in step D. "Maximum slope" was placed in dependent claim 7 and "maximum difference without fitting" in claim 9. So claim 1 is broad on the algorithm and comparatively narrow only on the acquisition protocol (A–B).

2. The cited prior art, characterized

US 4,592,089 (Bio Image Corp., 1986) — "Electrophoretogram analytical image processing system" (full text retrieved)

• Field: "image analyzer systems… for analyzing images of electrophoresis gels" — the same field.
• Electrophoretograms are stored as "digital images consisting of a matrix of pixel values" (→ element E's container; claim 10's array).
• Discusses silver staining of gels and quantifies spot intensity as "a useful measure of the quantity of protein present in the spot," summing "the absolute difference between the pixel values and the background level." Also references two-dimensional electrophoresis (Taylor "Human Protein Index"). → elements A, and the protein/2D context of claims 6, 11.

US 4,709,212 (GE, 1987) — "Method of enhancing image signal-to-noise ratio by combining NMR images of differing pulse sequence timing" (full text retrieved)

• Teaches acquiring multiple images of the same region at different acquisition timings (TE, TR), acknowledging that "simple addition… will not be simply interpretable," and instead fitting the pixel intensity to a mathematical function (an exponential in T₂/PD solved from the acquired intensities) and then computing a new image-intensity value I_c at each pixel for a selected reference timing TE_c.
• The stated object is an image with "improved signal-to-noise ratio." It explicitly contemplates that "a number of images at different values of TE_c could be generated, and then the signal-to-noise ratios… compared."
• → Teaches elements B¹ (multi-image acquisition of one region), C (per-pixel functional fit), D (compute value per a rule — here "at a reference time," which is literally claim 8), and E (generated/combined image) — but in NMR, not gel electrophoresis.

Fischer & Kochanski, "Optimal Addition of Images for Detection and Photometry," Astron. J. 107(2):802–810 (Feb. 1994) (full text/abstract retrieved)

• The examiner's sole non-patent citation. Explicitly motivated by "insufficient dynamic range" as a reason to "break up long CCD exposures into a series of shorter ones," then optimally co-add frames of a single field to maximize signal-to-noise.
• "Our techniques are strictly linear so that the images remain suitable for quantitative measurement and modeling." Discusses empirical fits for the weighting of individual frames.
• → Teaches the reason (dynamic-range extension) and the method (combine multiple frames of one field by a rule) underlying elements B/D/E — again outside the gel context.

US 5,571,643 (Martin, 1996) — "Spectrophotometric quantitation of images in X-ray film and electrophoresis gel bands" (full text retrieved)

• Quantitation of silver-stained electrophoresis gel bands and X-ray film; expressly frames the problem as measuring "the size and darkness of an image"; discusses densitometry creating "digital records of these samples" and comparison to a standard curve. → elements A; supports claims 4/5 context; corroborates the acknowledged problem of quantitating both faint and intense features.

Title-only references (provisional): US 4,909,920 (Helena, "Automatic electrophoresis apparatus and method," 1990); US 5,420,628 (RDF, "Video densitometer with determination of color composition," 1995); US 5,275,710 (Labintelligence, gel electrophoresis system w/ optical stage, 1994); US 5,470,710 (Univ. Utah, "Automated hybridization/imaging device for fluorescent multiplex DNA sequencing," 1995); US 5,795,716 (Chee, "Computer-aided visualization and analysis system for sequence evaluation," 1998). US 6,351,547 (GE, filed 1999-04-28, granted 2002-02-26) is §102(e) art but appears tangential (digital-image formatting to a communications standard).

Not prior art (post-dating the 1999-09-01 priority under the "Cited By" list): US 6,909,459 / US 2004/0036775 / WO 2004/019126 (Alpha Innotech, 2002) and US 7,155,050 / US 7,881,517 (GE Healthcare, priority 1999-11-03). These cannot be used; I note them only to exclude them.

Background admission (the patent's own specification): the Background states that with a single image "it is difficult if not impossible to measure accurately intensities of both faint and intense spots," that "no exposure time value is optimal for quantifying all objects… Multiple scans taken for different exposure times are required," and that for fluorescent gels "we multiply scans with different exposure and/or illumination settings are required." This is a §103-relevant admission that the problem and the multi-image approach were known.

3. Asserted obviousness combinations

Combination A — US 4,709,212 + US 4,592,089 (+ Bevington) — primary challenge to claim 1

• US 4,592,089 supplies elements A (stained gel), the pixel-matrix electrophoretogram, the protein-spot-intensity context, and the composite/array container (E / claim 10).
• US 4,709,212 supplies elements B–E in substance: acquiring several images of the same region at different acquisition timings, fitting per-pixel intensities to a mathematical function, computing an optimal intensity value per a predefined rule ("at a reference timing"), and inserting it into a generated composite image; its stated purpose is improving the image's information content (SNR).
• Bevington, Data Reduction for the Physical Sciences (McGraw-Hill, 1969), pp. 134–163 — the standard text the specification itself cites for "least-squares fit to a polynomial" — supplies the routine curve-fitting mathematics.

Motivation / why a PHOSITA would combine: The problem is identical and admitted — a single acquisition cannot capture both faint and intense objects, so multiple images of the same object must be combined. US 4,709,212 was issued precisely to solve "combining images acquired at different timing to yield one image with improved SNR," using per-pixel curve fitting. A PHOSITA confronting the admitted dynamic-range limitation of single gel images would look to the established image-processing art of combining co-registered frames by fitting a model to each pixel; applying that known combination technique to the known electrophoretogram pixel matrix is a predictable use of the prior art (KSR). The specification itself concedes the constituent techniques (multiple scans; digital pixel matrices; curve fitting via Bevington) were known; only the particular application is asserted as new.

Combination B — Fischer et al. + US 4,592,089 (or US 5,571,643) (+ Bevington) — dynamic-range rationale

• Fischer supplies the explicit motivation ("insufficient dynamic range") and a rule-based combination of multiple images of one field into a quantitative composite, plus empirical fits for weighting.
• US 4,592,089 / US 5,571,643 supply the gel/silver-stain/protein-quantitation context and the pixel-array representation (A, E).

Motivation: Fischer is the very reference the examiner used to show that co-adding multiple frames of a single field to overcome limited dynamic range and to preserve quantitative measurement was known. Transferring that technique from stellar photometry to gel densitometry is a predictable application to a known problem; dynamic range is a general property of photodetection, and a PHOSITA seeking "to quantify all objects simultaneously" (the patent's own articulation of the goal) would foresee success.

Combination C — Fischer + US 4,709,212 + US 4,592,089 — belt-and-suspenders

For a validity challenge, this triad covers every step with explicit text: the gel/pixel/silver-stain context (US 4,592,089), the per-pixel functional fit and reference-timing value computation (US 4,709,212), and the dynamic-range rationale and optimal multi-frame combination (Fischer). Motivation is supplied by the problem, the admitted background, and Fischer's express purpose.

4. Dependent claims 2–12

5. The likely non-obviousness strongholds (be candid)

1. "during said staining step" (element B). The retrieved art shows multi-image acquisition of the same object (US 4,709,212 — different timings; Fischer — different exposures/conditions), but I did not verify any reference teaching repeated scanning of a gel specifically while it is being stained/developed, to capture different stages of stain development. Fischer's frames are across "observing runs," not an in-progress biochemical development. The patent's own Background ties the acquisition to development ("spots… appear, darken, and ultimately reach saturation… no single point in time"). This limitation is the clearest structural distinction of claim 1, and it is where a validity challenge is weakest (and where infringement contentions would focus if the patent were live). A challenger would need art showing in-process monitoring during staining/development — best candidate among the cited set is US 4,909,920 ("Automatic electrophoresis apparatus") if its full text discloses iterative imaging during processing; I could not verify this from the present record.
2. Claim 7's "maximum slope" rule. Neither retrieved reference computes a maximum-slope point of a fitted development curve; US 4,709,212's rule is "value at a reference time." Absent art teaching slope-extremum selection of a developing signal, claim 7 is the most defensible. (Note claim 7 depends on claim 1, so it inherits the "during staining" limitation as well.)
3. No teaching-away that helps. US 4,709,212 warns that "simple addition… will not be simply interpretable" — but that caution endorses the fitting approach and thus supports, not defeats, the combination.

Secondary considerations (Graham factor 4): The specification contains no comparative or unexpected-results data; the claimed benefit (extended dynamic range) is the predicted result of combining multiple exposures. The Background articulates a known, long-felt need — which, if anything, weakens non-obviousness by showing the goal was recognized. No evidence of commercial success, copying, or licensing is on the record. On the present record, secondary considerations do not rescue claim 1 or claims 2–6 and 8–12.

6. Bottom line

• Claim 1 is, in my assessment, more likely than not obvious under §103 over US 4,709,212 + US 4,592,089 + Bevington, and alternatively over Fischer et al. + US 4,592,089/US 5,571,643 + Bevington — provided the challenger can meet the "scanning repeatedly during staining" limitation. Motivation is strong: an admitted, well-known dynamic-range problem; two analogous image-combining teachings (one per-pixel curve-fitting in US 4,709,212; one explicitly dynamic-range-driven in Fischer); and routine fitting mathematics (Bevington).
• Claims 8 and 10 are the most exposed (US 4,709,212 teaches reference-time computation; US 4,592,089 teaches pixel matrix/arrays).
• Claim 7 (and, to a lesser extent, the "during staining" acquisition step of claim 1) is the most defensible — no retrieved reference teaches selecting a maximum-slope value from a developing-signal curve.
• Claim 12 (nucleic acid) is supported only weakly because US 5,470,710 was verified by title only.

Explicit limitations of this analysis

• I analyzed only the references appearing in this patent's Prior Art section, as instructed. I did not run an independent prior-art search; the true invalidity picture could be materially stronger or weaker with art outside this list.
• Characterizations of US 4,909,920, US 5,420,628, US 5,275,710, US 5,470,710, and US 5,795,716 rest on titles/metadata only; these should be re-verified in full text before any of the above is relied on, especially for the "during staining" element and for claim 12.
• Claim text is quoted as rendered on the page, subject to the 2003-09-02 certificate of correction; the literal corrected claim language should be confirmed against the printed patent before any formal contention.
• This is a technical patentability/validity analysis, not a legal opinion, and it assumes pre-AIA §103 applies given the 1999 filing/priority date.