US 4939667

I'll search for authoritative USPTO and litigation/docket information on US 4939667, then reconcile it with the full patent text provided.

Now let me search specifically for litigation/docket activity and USPTO records.

Summary Report: US Patent 4,939,667

Verification note: I interpreted the identifier literally as US 4,939,667 (not a similar number such as 4,894,786, 4,937,667-adjacent, or the '966-family patents that appear in searches). The authoritative full text you supplied, plus Google Patents, Justia, and FreePatentsOnline, all corroborate the same document.

1. Bibliographic Data

2. Abstract (verbatim)

"A signal processing method for analyzing an autoradiograph by subjecting digital image data containing an autoradiograph of a resolved pattern to digital signal processing to obtain locational information on radioactively labeled substances in the form of numerals and/or symbols is disclosed. The resolved pattern is formed by resolving radioactively labeled substances, such as, for example, nucleic acids, in a one-dimensional direction on a support medium. The total boundary area of the resolved patterns throughout the digital image data is determined by first dividing the digital image data into two or more blocks along a direction perpendicular to the resolving direction and then preparing a waveform composed of positions along the direction perpendicular to the resolving direction for each block by compiling the digital image date for each block. The waveform is used to establish the boundary of the area of the resolved pattern for each block, and the boundaries of the resolved pattern area of each block are interpolated to establish the total boundary area of the resolved patterns. This method is an easily conducted, accurate analytical method for obtaining locational information and, thus, the identity of a wide variety of radioactively labeled substances."

3. Plain-Language Overview of the Independent Claims

There are four independent claims: claims 1, 7, 12, and 18. All four share the same core algorithmic idea; they split along two axes — (i) how the image was captured (stimulable phosphor sheet vs. radiographic film), and (ii) whether the method also isolates individual lanes ("resolved rows"). Every claim recites sub-steps (a)–(d) or (a)–(e).

Claim 1 — Phosphor-sheet capture; whole pattern boundary.
Analyzes a "one-dimensional" autoradiograph (e.g., a DNA sequencing gel) that was captured by exposing a stimulable phosphor sheet, then scanning it with stimulating rays and photoelectrically detecting the stimulated emission. The steps: (a) chop the digital image into two or more blocks along the axis perpendicular to the resolving direction; (b) collapse each block into a one-dimensional waveform (a "projection") along that perpendicular axis; (c) use that waveform to find the edge/boundary of the resolved pattern in each block; (d) interpolate the per-block boundaries to get the overall boundary of the whole pattern.

Claim 7 — Film capture; whole pattern boundary.
Same (a)–(d) algorithm, but the input is an autoradiograph recorded on an X-ray (radiographic) film and read out photoelectrically. It is essentially the film-based twin of claim 1.

Claim 12 — Phosphor-sheet capture; pattern boundary plus individual lane boundaries.
Same phosphor-sheet capture as claim 1, but adds two steps: (d) use the waveform to find the boundary of each individual lane (resolved row) within each block, and (e) interpolate both the pattern boundaries and the lane boundaries so that each lane is traced across the entire image. This is the "draw out each lane" version.

Claim 18 — Film capture; pattern boundary plus lane boundaries.
The film-based twin of claim 12 — same pattern-plus-lane algorithm on a radiographic-film image.

Dependent claims (2–6, 8–11, 13–17, 19–23) add implementation detail, notably:

• Compiling all pixels along the resolving direction (e.g., claims 2, 8, 13, 19) vs. sampling at equal intervals and compiling only the extracted subset (claims 3, 9, 14, 20).
• Setting the pattern boundary via a mean signal level → threshold on the projection (claims 4, 10, 15, 21).
• Locating the lane boundary at the minimum signal level (sign-of-difference inversion) inside the pattern (claims 16, 22).
• Determining boundaries on a central line of each block and joining corresponding points to build the full boundary (claims 5, 11, 17, 23).
• The "utility" dependent claim 6: the labeled substances are nucleic acids/derivatives/cleavage or synthetic products, and the locational information is base-sequence information.

Technical gist in one sentence: Rather than trying to trace a curved or slanted electrophoresis pattern globally, the method slices the image into blocks, projects each block to a 1-D profile, finds edges and lane minima locally, then interpolates the local landmark points into smooth global boundary lines — which works even when lanes are curved or skewed.

4. Prior Art Cited (front page) — 8 references, all relevant context

• US 4,617,468 — Stimulable phosphor sheet with hydrophilic surface (Fuji)
• US 4,629,891 — Radiation image read-out method and apparatus (Fuji)
• US 4,665,312 — Signal detecting method in autoradiography (Fuji)
• US 4,706,192 — Input apparatus for entering base sequence information of the gene (Hitachi Software Engineering)
• US 4,720,786 — Method of compensating for offset distortion in rows of electrophoretic patterns (Fuji)
• US 4,734,581 — Autoradiographic process (Fuji)
• US 4,748,326 — Radiation image read-out apparatus (Fuji)
• US 4,777,597 — Signal processing method in autoradiography (Fuji)

The patent also relies on several co-pending Fuji applications (Ser. Nos. 664,405; 837,037; 568,877; 730,034; 91; 917,609), all recited as the source of the phosphor-sheet autoradiography technique.

Notable later citation: US 11,133,087 B2 (Li-Cor, Inc., granted 2021) — "Adaptive lane detection systems and methods" — which is a modern descendant of exactly this lane-detection problem.

5. USPTO / Docket / Litigation Search — Result and Uncertainty

USPTO records: The patent is confirmed as granted 1990-07-03, assigned to Fuji Photo Film Co., Ltd., with maintenance fees paid at the 4-year, 8-year, and 12-year intervals (last recorded FPAY 2001-09-19). Status is Expired – Lifetime, with an anticipated expiration of 2007-12-28. There is no indication in the record of any reissue, reexamination, disclaimer, or adverse PTAB proceeding.

CAFC 2026 dockets: I searched specifically for "4939667" combined with Federal Circuit / CAFC / IPR / PTAB / litigation terms. I found no 2026 (or any) CAFC docket, appeal, or ITC action involving US 4,939,667. My honest assessment: this is expected — the patent expired in 2007, and its 20-year term ran out before the modern AIA post-grant review era. Any search hits tying "Fuji Photo Film" to 2026 Federal Circuit or UPC activity (e.g., the Fujifilm v. Kodak UPC litigation over EP 3 511 174, or the Jazz Photo/Benun appeals) concern entirely different patents and different technologies and must not be attributed to US 4,939,667.

Explicit uncertainty flags:

1. I could not retrieve a live USPTO PatentCenter or PTAB "Patent Trial and Appeal Board" record page directly; my USPTO-side facts come from the Google Patents front-matter (which itself notes its legal-status listings are assumptions, not legal conclusions).
2. I cannot rule out a purely defensive, never-docketed licensing use of the patent by Fuji in the 1990s–2000s. Absent a docket record, I will not assert one.
3. The claims as reproduced contain several typographical errors carried in the original printing ("meduim," "positionl," "total boundary are of the resolved patterns," "stimilated emission," "using the waveform the establish"). I have not auto-corrected these; they are reproduced as literal claim language, and any infringement/validity analysis should quote them exactly as issued.

6. Bottom Line

US 4,939,667 is a 1990 Fuji Photo Film patent (Hara & Sato) covering a block-based lane/pattern boundary detection algorithm for autoradiographs, aimed at automated DNA base-sequence reading from stimulable-phosphor or X-ray-film images. It is long expired (2007) and, based on my searches, has no CAFC 2026 docket presence and no recorded post-grant proceedings. Its independent claims 1 and 7 cover whole-pattern boundary extraction (phosphor vs. film variants), and claims 12 and 18 extend that to per-lane boundary extraction.

I'll search for litigation involving US patent 4939667.

Initial results are mostly noise. Let me search more targeted sources.

Let me check litigation-specific databases directly.

Litigation Search Report — U.S. Patent No. 4,939,667

Prepared: April 26, 2026
Subject patent: U.S. Patent No. 4,939,667 ("Signal processing method for analyzing autoradiograph"), granted July 3, 1990; application 07/138,395 filed December 28, 1987; priority JP 61-309754 (1986‑12‑27); original assignee Fuji Photo Film Co., Ltd.; current listed assignee Fujifilm Holdings Corp.

Bottom line

I found no known litigation (district court, ITC, or Federal Circuit) involving U.S. Patent No. 4,939,667, and no PTAB post-grant proceeding (IPR/PGR/CBM) against it.

I want to be explicit about the limits of this answer: I searched patent-litigation aggregators and general web sources (including the Unified Patents litigation/PTAB caselist pages referenced in my instructions, Docket Alarm, CourtListener, FreePatentsOnline, Google Patents, and Justia). The searches for the literal string "4939667" and "4,939,667" returned only non-litigation or unrelated hits, described below. I was unable to retrieve a positive confirmation from any source that this specific patent was ever asserted in a case, and I did not find any case caption naming it. I cannot rule out a case that is not indexed in the sources I could reach, so treat this as "no litigation known to me" rather than an absolute negative.

What the searches did return (and why it is NOT litigation for this patent)

To avoid any ambiguity, here are the hits for the literal number "4939667" that I evaluated and excluded because they do not concern U.S. Patent 4,939,667:

None of these is a patent-infringement action, ITC investigation, or appeal involving U.S. 4,939,667.

Verification of the patent's status (relevant to litigation likelihood)

• Legal status: Expired – Lifetime. The record shows an anticipated expiration date of December 28, 2007 (20 years from the December 28, 1987 filing date; the maintenance-fee "Year of fee payment: 12" was paid September 19, 2001).
• The patent family record lists only one granted U.S. patent and its Japanese counterpart JPS63167290A (and JP grant JPH0570792B2). No continuations, reissues, or reexaminations appear in the record.
• No post-grant proceedings (IPR/PGR/CBM) appear in the sources checked.
• The "Citations (8)" and "Cited By" tables on Google Patents list only prior-art citations and forward-citing documents — these are references, not litigation. Several cited patents (e.g., US 4,777,597; US 4,666,312; US 4,720,786; US 4,738,321; US 4,660,538 family members) are Fuji Photo Film sibling patents, which are frequently confused with the subject patent in casual searches; they are separate patents and were not the subject of any litigation I found either.

Why the absence of litigation is plausible (contextual note)

U.S. 4,939,667 is a data-processing/method patent covering a technique for detecting and drawing out resolved-pattern and "lane" boundaries from digitized autoradiograph image data (Claims 1–23). It claims a signal-processing method (not a physical product), it was owned by Fuji Photo Film, and it expired in 2007. Patents of this profile — internal image-analysis methods tied to a commercial instrument/system, never separately monetized — are rarely the subject of publicly reported infringement suits. That is consistent with, though not proof of, the absence of litigation.

Recommended next steps to obtain a definitive negative

If a formal clearance/freedom-to-operate or due-diligence negative is required, I recommend confirming against the subscription databases I could not fully query in this session:

1. PACER / Docket Alarm / Lex Machina — docket search on patent number 4,939,667 and on application 07/138,395.
2. Unified Patents Portal litigation caselist and PTAB caselist (filter by patent number).
3. Westlaw KeyCite / Lexis CourtLink — the same "Litigation Search Report" methodology the USPTO Central Reexamination Unit uses (KeyCite history + CourtLink open/closed dockets), which reliably surfaces any assertion for a given patent number.
4. USITC EDIS — to confirm no Section 337 investigation named this patent (Fuji Photo Film was an active ITC complainant in the 1998–2004 disposable-camera era, but on entirely different patents — e.g., Re. 34,168; US 4,884,087; US 4,833,495; US 4,972,649 — none of which is 4,939,667).

Summary table

Conclusion: Based on the searches performed, there is no known litigation involving U.S. Patent No. 4,939,667. All literal "4939667" hits were either unrelated patent numbers in other jurisdictions or non-patent records. I could not reach confirmatory results from every subscription litigation database, so this conclusion is subject to verification through PACER/Docket Alarm/Lex Machina and Westlaw KeyCite/Lexis CourtLink before being relied upon as a definitive negative.

I'll verify the ODP-derived "no proceedings" default against public sources before writing.

Let me run a couple more targeted checks for any proceeding or litigation naming this exact patent.

Proceedings overview

Zero AIA trial proceedings — the USPTO Open Data Portal returns no IPR, PGR, or CBM petition, institution decision, Final Written Decision, or appeal for US 4,939,667, and independent public-source searching (PTAB decisions, PTAB E2E docket materials, Federal Circuit opinions, CourtListener, and litigation reporting) surfaces no PTAB activity whatsoever — so the count is 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution-denied. The bottom line for a defendant is not "the patent is hardened" but the opposite: every one of the 23 claims is untested, and the patent term expired on 2007-12-28, so a demand letter citing US 4,939,667 today is not a validity fight at all — it is a statute-of-limitations problem for the sender, and the correct response is to demand proof of a live, unexpired right rather than to file an IPR.

No proceedings on file

• Type: N/A — no Inter Partes Review, Post-Grant Review, or Covered Business Method review has ever been filed against this patent.
• Filed: N/A
• Status: Verbatim from the structured data: "The USPTO ODP API returns no AIA trial proceedings for this patent as of the most recent ingest." Plain-English gloss: nothing was filed, so nothing was instituted, tried, terminated, or appealed.
• Judge panel: N/A
• Petition grounds: N/A
• Institution decision: N/A
• Final Written Decision: N/A — no claim of US 4,939,667 has ever been canceled, confirmed, or construed by the Board.
• Settlement / termination: N/A
• Appeal: N/A — no Board decision exists to appeal, and I found no Federal Circuit or CourtListener docket naming this patent.
• Defensive value: Neutral-to-positive in the sense that no adverse or preclusive PTAB record exists, but that is cold comfort either way: the patent's entire 20-year term ran from the 1987-12-28 filing date to 2007-12-28 (the structured record lists "Anticipated expiration — 2007-12-28" and legal status "Expired - Lifetime"), which means the § 286 six-year damages lookback closed on or about 2013-12-28 and no infringement claim of any kind can now be maintained on this patent. There is nothing left to invalidate and nothing left to litigate.

Disambiguation — near-misses that are not this patent

I checked these deliberately so they are not mistaken for proceedings on 4,939,667:

• IPR2019-00683 (ClearOne v. Shure) and PGR2020-00079 concern US 9,565,493 — a Shure microphone-array patent, unrelated.
• IPR2019-01094 (AMP Plus d/b/a ELCO Lighting v. DMF, affirmed in part at AMP Plus, Inc. v. DMF, Inc., No. 2023-1997 (Fed. Cir. 2025-03-19)) concerns US 9,964,266.
• An IPR petition referencing a "'493 patent" challenged as "Claims 1–14" over Misawa and Parulski references is a camera/imaging patent, not the Hara/Sato autoradiograph patent — the asserted art families (Misawa, Parulski, Watanabe, JP-A-11-355665) post-date this patent by more than a decade.
• Fuji Photo Film Co. v. Jazz Photo Corp., 394 F.3d 1368 (Fed. Cir. 2005), the ITC lens-fitted film package investigation (Inv. No. 337-TA-406), and Jazz Photo Corp. v. Int'l Trade Comm'n, 264 F.3d 1094 (Fed. Cir. 2001) assert disposable-camera patents (US 4,833,495; US 4,885,774; US 4,884,087; US 4,972,649; etc.). US 4,939,667 appears nowhere in those asserted-patent lists.

Strategic summary

Claim status. All 23 claims of US 4,939,667 — independent claims 1, 7, 12, and 18, plus dependents 2–6, 8–11, and 13–17 and 19–23 — are untested: none canceled, none sustained by any adjudicative body, none amended by reissue or reexamination certificate on the record. The only narrowing events I can find are ordinary prosecution events, not post-grant ones. The claims split into two pre-AIA families: claims 1–6 and 12–17 are tied to the stimulable phosphor sheet readout (irradiating with stimulating rays, photoelectrically detecting stimulated emission), and claims 7–11 and 18–23 are tied to the radiographic-film readout (visible image photoelectrically read out). Claim 6 is the base-sequence-specific dependent claim. Nothing here has been construed by the PTAB or any district court that I can locate.

Estoppel landscape. There is none, and that is the point. Because no IPR/PGR was ever instituted, § 315(e)(2) estoppel never attached to anyone — no petitioner, real party in interest, or privy is barred from any ground, and no prior-art ground is off the table by estoppel. But estopped or not, the grounds are irrelevant: with the term ended 2007-12-28 and the § 286 lookback closed around 2013-12-28, there is no live infringement theory to defend, and validity is moot absent some collateral use of the patent (e.g., as prior art, or in an ownership/licensing dispute over the 1987–2007 period). Practically, a defendant's invalidity position is: "Expired. Nothing to construe, nothing to invalidate, no damages window."

Pattern signals. No pattern at all. There is no repeat petitioner, no defensive aggregator (no Unified Patents, RPX, or similar entity) in the chain, no PTAB appeal history, and no Board certificate. The assignee chain is single: Fuji Photo Film Co., Ltd. (original assignee, per the 1987-12-25 assignment of Hara and Sato, Reel/Frame 004825/0838) → Fujifilm Holdings Corp. (current assignee per Google Patents). The patent's continued relevance is as cited prior art, not as an asserted right — notably it is cited by US 11,133,087 (Li-Cor, Inc., "Adaptive lane detection systems and methods"), a 2021 patent that is a remote descendant of this 1990 disclosure in the lane-detection art. That citation is a signal that the technology line survived, not that this patent is alive.

Recommended next steps

• If you are a defendant and this patent is in a demand letter, treat it as a paper tiger. There is no Final Written Decision to link to and no claim to quote, because no proceeding exists. The dispositive citation is the face of the patent and the structured record: priority 1986-12-27 (JP 61-309754), US filing 1987-12-28, granted 1990-07-03, anticipated expiration 2007-12-28, status "Expired - Lifetime" (US4939667A on Google Patents). Under 35 U.S.C. § 154(a)(2) and § 286, that puts every possible damages period outside the six-year lookback as of 2026-09-27.
• Do not file an IPR — it would be a waste of fees and would likely be denied. The Board has little reason to institute on a patent with no live enforcement, and the practical value is nil. If you nonetheless need to confirm the docket, search PTAB E2E directly at https://ptacts.uspto.gov/ptabweb and PATENTS End-to-End at https://patentsendtoend.uspto.gov for application number 07/138,395 — I found no proceeding there or in any indexed decision.
• Screen for a live relative before you dismiss the letter. Family members to check are the Japanese counterpart JP S63167290A / JP H0570792B2 and contemporaneous Fuji applications in the same program: US 4,777,597; US 4,801,?? / US 4,892,?? series; specifically the siblings named in the specification and the "Similar Documents" list — US 4,777,597, US 4,894,786, US 4,981,092-adjacent filings, US 4,972,325, US 4,958,281, US 4,980,827, and US 4,837,687 — all of which are also long expired, as are the cited US 4,617,468, US 4,629,891, US 4,665,312, US 4,706,192, US 4,720,786, US 4,734,581, and US 4,748,326. If the demand letter is actually pointing at a modern Fujifilm patent (e.g., Li-Cor-style lane-detection or sequencing-imaging claims), this patent's real role is as an anticipation/obviousness reference against that later patent, not as a sword.
• One record caveat, flagged rather than smoothed over: the public legal-events record for this patent shows maintenance-fee payments for years 4 (1994-01-03), 8 (1998-01-05), and 12 (2001-09-19) and does not show a year-16 fee payment, which would have been due around 2005. Google's legal-events data for 1980s–1990s patents is frequently incomplete, so I cannot tell you with confidence whether the patent lapsed for non-payment before the 2007-12-28 term date or simply expired by term. Either way the conclusion is identical for a defendant today — the rights are gone — but if precise expiration timing ever matters (e.g., a pre-2007 damages tail), obtain the certified maintenance-fee history from the USPTO Patent Center rather than relying on the aggregated record.
• Bottom line. No PTAB activity exists on US 4,939,667, and the absence is not the interesting signal here — the expiration is. Well-asserted patents attract IPRs; this one never attracted one because Fuji's assertion campaigns of that era (the lens-fitted film package wars) were built on its camera patents, not its autoradiography signal-processing portfolio. If you receive a demand on this patent number, the letter is either a clerical error or an attempt to leverage an expired right; respond with the expiration date and decline to engage on validity.

I'll verify the assignment record and surrounding facts before drafting.

Let me check whether the 2007 Fujifilm bulk assignment (reel 018904/0001) covers this specific patent.

US 4,939,667 — Assignment / Ownership Reconstruction

"Signal processing method for analyzing autoradiograph" · App. 07/138,395 · Filed 1987-12-28 · Granted 1990-07-03 · JP priority 1986-12-27 (JP61-309754)

Inventors

Both inventors list their residence as Minami-ashigara, the site of Fuji Photo Film's registered headquarters and principal Kanagawa plant (210 Nakanuma, Minami-Ashigara-shi). Both executed an "ASSIGNMENT OF ASSIGNORS INTEREST" in favor of Fuji Photo Film Co., Ltd. before the US filing date (effective 1987-12-25, application filed 1987-12-28), which is the standard profile of employee-inventors under a pre-existing obligation to assign. I did not find any independent employment record confirming they were Fuji Photo Film employees, and no inventor-departure data is available for either — there is no evidence of inventors leaving the assignee, and no successor entity in the chain suggests a portfolio break-up.

Unusual patterns: none detected. This is a routine single-assignee corporate filing, and both inventors remained associated with the patent family through the assignee (the specification cross-references several co-pending Fuji Photo Film applications: Ser. Nos. 664,405; 837,037; 568,877; 730,034; 91; 917,609 — all Fuji internal program filings).

Original assignee

Fuji Photo Film Co., Ltd., 210 Nakanuma, Minami-Ashigara-shi, Kanagawa, Japan (a Corp. of Japan per the recorded assignment).

• Primary line of business at issuance (1990): photographic films, photographic paper, imaging chemistry, and — critically for this patent — the radiation image recording and reproducing ("computed radiography") business, which Fuji built around the BaFBr:Eu²⁺ stimulable phosphor sheet recited in claims 1 and 12.
• Did they ship a product embodying the claims? Effectively yes, at the systems level. The patent claims a signal-processing method for extracting lane/pattern boundaries from autoradiograph image data read out of a stimulable phosphor sheet or X-ray film. Fuji commercialized this exact stack as the FCR (Fuji Computed Radiography) product line and as the autoradiography bio-analysis scanners that used stimulable phosphor imaging plates. This is an operating-company, product-anchored patent — claims 1 and 12 are expressly tied to reading the plate on a Fuji stimulable phosphor read-out apparatus.
• Current status: Operating. Fuji Photo Film Co., Ltd. changed its corporate name to FUJIFILM Holdings Corporation effective 2006-10-01 when it adopted a holding-company structure, and the operating businesses were succeeded by FUJIFILM Corporation (see timeline below). Fujifilm remains a publicly traded, solvent multinational (TYO: 4901) with ~JP¥2.3T revenue. No bankruptcy, receivership, or assignment-for-the-benefit-of-creditors event exists anywhere in this chain.

Assignment timeline

Only one assignment is recorded against US 4,939,667 in the USPTO assignment record as surfaced via Google Patents' legal-events feed for this patent. The full chain is a single pre-filing inventor→corporate assignment, followed by the company's own corporate-name/holding-structure evolution.

1. 1987-12-25 (executed) / recorded 1987-12-28 — Reel 004825/0838

• Conveyance: Assignment ("ASSIGNMENT OF ASSIGNORS INTEREST")
• Assignor: Hara, Makoto; Sato, Shu (both individually)
• Assignee: Fuji Photo Film Co., Ltd., a Corp. of Japan, No. 210 Nakanuma, Minami-Ashigara, Kanagawa, Japan
• Correspondent: Not exposed in the retrieved record. The Assignment Center/Google Patents legal-events entry for reel 004825/0838 shows only the assignors, assignee, and reel/frame — no recording attorney or correspondent firm is surfaced. I therefore cannot assess recurrence of a recording correspondent, and I will not substitute the prosecution firm for it (see Note below).
• Context: Pre-filing in-house capture of employee inventions — the assignment was executed three days before the US application was filed and recorded the same day the application was filed. Standard operating-company practice, not a transfer of any kind.
• Family counterpart: The same invention was filed in Japan as JP61-309754 (1986-12-27) → published JPS63167290A (1988-07-11) → granted JPH0570792B2 (1993-10-05), assigned to the same company.

2. 2007-01-30 (effective) / recorded 2007-02-15 — Reel 018904/0001 — applicability to this patent NOT confirmed

• Conveyance: Assignment
• Assignor: FUJIFILM HOLDINGS CORPORATION (formerly Fuji Photo Film Co., Ltd.)
• Assignee: FUJIFILM CORPORATION, Japan
• Correspondent: Not exposed in the INPADOC legal-status text that reports this reel/frame.
• Context: Internal corporate reorganization only — this is the documented post-2006-holding-company migration of the Fujifilm US patent estate from the renamed holding company down to the operating entity.
• ⚠️ Evidence caveat: This reels/frame appears in INPADOC legal-status records for other Fujifilm US patents (e.g. US 2006/0152562 A1's legal-status history reports "NEW OWNER: FUJIFILM CORPORATION, JAPAN; ASSIGNOR: FUJIFILM HOLDINGS CORPORATION (FORMERLY FUJI PHOTO FILM CO., LTD.); REEL/FRAME:018904/0001; EFFECTIVE DATE 20070130"). It was a bulk recording across a large Fujifilm portfolio, so it plausibly covers the '667 patent. However, Google Patents' legal-events feed for US 4,939,667 does not list it, and Google still renders the "Current Assignee" as Fujifilm Holdings Corp — which is the pre-2007 holder name. I could not open the Assignment Center record for reel 018904/0001 to confirm the patent list. Treat link #2 as probable but unverified for this specific patent.

3. Termination — 2007-12-28

• Anticipated expiration reached (20 years from the 1987-12-28 US filing date). Google Patents status: "Expired – Lifetime." Maintenance fees were paid at years 4 (1994-01-03), 8 (1998-01-05), and 12 (2001-09-19); the estate then simply ran to term. No post-expiration transfer is possible or recorded.

Note on the prosecution firm — not an assignment correspondent. The face of the patent names law firm Sixbey, Friedman, Leedom & Ferguson as prosecution counsel (per Justia's patent history for 4,939,667; primary examiner Kevin J. Teska). That is prosecuting counsel, not the correspondent of record on any assignment filing, and it must not be counted as a "repeat correspondent across the chain" signal.

Timeline diagram

timeline
    title Ownership of US 4939667
    1986 : JP priority application filed
    1987 : US application filed 28 Dec
         : Inventors assign to Fuji Photo Film
         : Recorded reel 004825 frame 0838
    1990 : US patent 4939667 granted
    1994 : Maintenance fee paid year 4
    1998 : Maintenance fee paid year 8
    2001 : Maintenance fee paid year 12
    2006 : Fuji Photo Film becomes FUJIFILM Holdings
    2007 : Holdings to Fujifilm Corp support
         : Reel 018904 0001 not confirmed here
    2007 : Patent term ends 28 Dec

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. The only recorded assignee in the entire life of the patent is Fuji Photo Film Co., Ltd. (reel 004825/0838), an operating photographic/imaging manufacturer. There is no "IP / Patents / Licensing / Holdings / Ventures" successor, no Delaware or Texas single-purpose LLC, no registered-agent-service address anywhere in the chain.
2. Known asserter in the chain — NOT PRESENT. No assignee or assignor matches any entity on the supplied list (Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities). The forward citations to this patent (Hyseq 5,972,619; Affymetrix 6,242,180 and family; GE 6,201,890; University of Utah 6,208,941; Jydsk Telefon WO92/05427; Li-Cor 11,133,087) are citations, not ownership transfers — none of them acquired the '667 patent.
3. Repeat correspondent across the chain — NOT PRESENT / NOT ASSESSABLE. There is exactly one recorded link, so recurrence is definitionally impossible. Additionally, the correspondent field was not retrievable for reel 004825/0838. No NPE-recording attorney name is associated with this patent at all. Do not read the prosecuting firm Sixbey, Friedman, Leedom & Ferguson as a correspondent finding.
4. Cascading transfers — NOT PRESENT. Zero transfers in 24 months; the only movement in the patent's 20-year life is a single pre-filing inventor assignment plus (probably) one intra-group reorganization 19 years later.
5. Pre-litigation transfer — NOT PRESENT. No infringement suit naming US 4,939,667 was found in any of the searched sources; there is no pre-suit assignment to pair with a filing date. (Fujifilm is an active plaintiff in other disputes — ITC Inv. No. 1076 vs. Sony over magnetic tape, Fujifilm v. Kodak UPC_CFI_359/2023, PTAB IPRs against Sony — but none of those assert this patent.)
6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 event. Fuji Photo Film → FUJIFILM Holdings (2006-10-01) was a solvent holding-company reorganization, and Fujifilm remains a going concern. This is not a Kodak/Nortel/Polaroid fact pattern.
7. Privateering — NOT PRESENT. No transfer from Fujifilm to any NPE asserting on its behalf. Fujifilm asserts its own patents in its own name (e.g. ITC complainant "Fujifilm Corporation" with "Fujifilm USA"; Fujifilm v. Benun, Fed. Cir.). The patents cited against Fujifilm in this space run the other direction — e.g. Optimum Imaging Technologies LLC v. Fujifilm, E.D. Tex. 4:23-cv-00922 — i.e. Fujifilm is the NPE target, not the source.
8. Defensive aggregator — NOT PRESENT. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates by expiry in the hands of the original corporate family. The patent was never neutralized by aggregation; it was simply run to term.

Verdict

Insufficient data — meaning: only the original assignment exists.

The single recorded assignment is reel 004825/0838, executed 1987-12-25 and recorded 1987-12-28, conveying Hara and Sato's rights to Fuji Photo Film Co., Ltd. No licensing-only LLC, no known asserter, and no NPE-side correspondent appears anywhere in the record. There is a probable 2007-01-30 intra-group transfer recorded at reel 018904/0001 (FUJIFILM HOLDINGS CORPORATION → FUJIFILM CORPORATION) that is documented on sibling Fujifilm US patents but is not confirmed for this specific patent because Google's legal-events feed for US 4,939,667 lists no 2007 assignment and still names Fujifilm Holdings Corp as current assignee. Every NPE signal above resolves to not present; the only ownership movement is a corporate name change and holding-company reorganization at a solvent, publicly traded operating company. The patent expired for failure to attain term on 2007-12-28 and is unassertable today.

Verify:

• USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search patent 4,939,667; then pull reel 004825/0838 and cross-check whether reel 018904/0001 enumerates this patent)
• Google Patents legal events — https://patents.google.com/patent/US4939667/en
• JP family — JPS63167290A / JPH0570792B2

Confidence and limits: High confidence that no NPE, shell-entity, or cascading-transfer chain exists for this patent. Explicitly lower confidence on two data points: (a) the assignment correspondent for reel 004825/0838 was not retrievable from the sources available to me, so the repeat-correspondent signal could not be tested empirically; and (b) reel 018904/0001's coverage of this patent is inferred from sibling patents, not verified against the record itself.

I'll verify the patent and each cited reference, then assess relevance to the claims.

Let me gather details on the remaining cited references.

Prior-Art Analysis — U.S. Patent No. 4,939,667

Prepared: April 26, 2026 | Source of record: full text of US 4,939,667 (Google Patents), including the "Patent Citations (8)" table, which is treated as authoritative for citations and dates.

Verification note (strict-identifier rule): The number searched was literally 4,939,667. It resolves to "Signal processing method for analyzing autoradiograph," application 07/138,395, filed Dec. 28, 1987, priority JP 61‑309754 (Dec. 27, 1986), granted July 3, 1990, inventors Makoto Hara and Shu Sato, original assignee Fuji Photo Film Co., Ltd., 23 claims. No other patent was substituted in this analysis.

1. The patent being analyzed (recap for cross-reference)

The single point of novelty common to all four independent claims is the combination of: (a) dividing the digital image data into two or more blocks along a direction perpendicular to the resolving direction; (b) compiling the data per block into a waveform (a projection) along that perpendicular direction; (c) establishing a resolved-pattern boundary per block; and (d) interpolating across blocks to establish the total boundary (claims 12/18 add per-block resolved-row boundaries and the row-boundary interpolation). Claims 1 and 12 recite the stimulable-phosphor arm; claims 7 and 18 recite the radiographic-film arm.

Legal framework. Because the application was filed Dec. 28, 1987, the pre‑AIA version of 35 U.S.C. § 102 governs. Mapping to the general Dec. 28, 1986 / Dec. 28, 1987 critical dates:

• § 102(b) — reference patented more than one year before the U.S. filing date (i.e., on/before Dec. 28, 1986).
• § 102(a) — reference patented before the applicant's invention date.
• § 102(e) — U.S. patent granted on an application filed before the applicant's invention (all eight cited documents satisfy this, as their applications were all filed in 1983–1985).

A single reference anticipates only if it discloses every element of a claim, arranged as claimed. As shown below, none of the eight references discloses the block-division/per-block-projection/interpolation combination, so none is an anticipation reference for any independent claim; they function as § 102(e) background art and as § 103 obviousness art.

2. Reference-by-reference analysis

2.1 U.S. Patent No. 4,617,468

"Stimulable phosphor sheet with hydrophilic surface" — Shiraishi et al.; Fuji Photo Film Co., Ltd.
· Priority Feb. 24, 1983 · Issued Oct. 14, 1986 · § 102(b) + § 102(e)

• Description: Directed to the recording medium itself — a stimulable phosphor sheet given a hydrophilic surface treatment (e.g., for handling aqueous/gel samples). It concerns sheet structure/materials, not image-data processing.
• Claim relevance: Touches only the "stimulable phosphor sheet" element recited in the preambles of claims 1 and 12 (and, by contrast, is the medium that claims 7/18 replace with a radiographic film).
• § 102 verdict: No anticipation of any claim. It discloses no dividing, compiling, boundary-establishing, or interpolating steps. Purely medium-level background art.
• Caveat: full claim text of this reference was not retrievable on the search budget used; the above rests on the title/field and the patent's own citation entry.

2.2 U.S. Patent No. 4,629,891

"Radiation image read-out method and apparatus" — Nakajima et al.; Fuji Photo Film Co., Ltd.
· Priority Sep. 13, 1984 · Issued Dec. 16, 1986 · § 102(b) + § 102(e)

• Description: Read-out hardware/method for scanning a stimulable phosphor sheet with stimulating rays and photoelectrically detecting the stimulated emission to produce an image signal — the read-out infrastructure underlying claim 1/12's "irradiating said phosphor sheet with stimulating rays and photoelectrically detecting the autoradiograph as stimulated emission."
• Claim relevance: Only the read-out clause of claims 1/12. No data-processing algorithm.
• § 102 verdict: No anticipation of any claim. Background/support art for the phosphor read-out step.
• Caveat: same retrieval limitation as 2.1.

2.3 U.S. Patent No. 4,665,312 — among the two most relevant

"Signal detecting method in autoradiography" — Shiraishi et al.; Fuji Photo Film Co., Ltd.
· Priority Mar. 31, 1983 · Issued May 12, 1987 · § 102(a)/§ 102(e)

• Description (substantially retrieved): Applies to an autoradiograph on a stimulable phosphor sheet (Claim 1 thereof expressly covers the phosphor arm, and a sibling — U.S. 4,884,200 — covers the film arm). The method (1) scans the sheet at at least two different positions that traverse the one-dimensional distribution, detects distribution points for each scan, and prepares a continuous line (straight, polygonal, or curved) along the corresponding distribution points to assign the one-dimensional distribution direction; then (2) scans along that direction to obtain the digital signal.
• Claim relevance: This is the closest cited art on the "interpolating … points" idea — it joins detected points into a continuous line. However, the line it fits is the scanning (resolving-axis) direction line through band/spot centers, not boundaries of a resolved pattern area or of lanes, and it contains no division into blocks perpendicular to the resolving direction and no per-block projection waveform.
• § 102 verdict: No anticipation of claims 1, 7, 12, or 18 (and therefore none of the dependents). It lacks the block-division and per-block compile/boundary features. Best characterized as § 102(e) / § 103 art that a challenger would combine with a boundary-detection reference.

2.4 U.S. Patent No. 4,706,192

"Input apparatus for entering base sequence information of the gene" — Nasu et al.; Hitachi Software Engineering Co., Ltd.
· Priority Apr. 27, 1984 · Issued Nov. 10, 1987 · § 102(a)/§ 102(e)

• Description (retrieved): A manual data-entry apparatus — a digitizer/keyboard (with a scale/magnifier and a speech-synthesis output) by which an operator traces band positions on an X-ray film and the computer converts coordinates to base codes A/G/C/T (Maxam–Gilbert). Significant for showing computer-assisted DNA sequencing, but it is a human-driven input device, not an automated boundary-detecting signal-processing method.
• Claim relevance: Only marginal background for claim 6-type subject matter (nucleic acids / base sequence). It neither digitizes the image automatically nor performs block division, projection, or interpolation.
• § 102 verdict: No anticipation of any claim. Different technical character (manual input hardware).

2.5 U.S. Patent No. 4,734,581

"Autoradiographic process" — Hashiue; Fuji Photo Film Co., Ltd.
· Priority May 2, 1984 · Issued Mar. 29, 1988 · § 102(e)

• Description: An autoradiographic process using the stimulable-phosphor (radiation-image recording and reproducing) technique in place of X-ray film — i.e., the general recording workflow invoked in the preamble of claims 1/12.
• Claim relevance: Preamble-level only (recording the autoradiograph on the phosphor sheet).
• § 102 verdict: No anticipation of any claim. Processing-algorithm features absent.
• Caveat: claim text not retrieved; characterization rests on title/field and the patent's citation entry.

2.6 U.S. Patent No. 4,720,786 — notable: same inventor

"Method of compensating for offset distortion in rows of electrophoretic patterns" — Hara; Fuji Photo Film Co., Ltd.
· Priority Apr. 19, 1985 · Issued Jan. 19, 1988 · § 102(e) (see "by another" note)

• Description (retrieved): Operates on digital signals from an autoradiograph of plural resolved rows ("lanes") of base-specific DNA/RNA fragments. It (1) detects at least two bands in the lower part of each row and numbers them; (2) obtains a correlation (regression line/curve) between band number and migration distance per row; and (3) determines the difference in migration distance between rows and corrects band positions for "offset/smiling" distortion to establish base sequence. Acknowledges both the phosphor-sheet and radiosensitive-film sources.
• Claim relevance: Directly concerns resolved rows (lanes) (relevant to claims 12/18's "plurality of resolved rows"), works on autoradiograph digital signals, and deals with lane-to-lane position variation — conceptually adjacent to the subject patent's need to track lanes that are curved or inclined.
• § 102 verdict: No anticipation of claims 1, 7, 12, or 18. It corrects band positions along the resolving direction via regression on band number; it does not divide the image into blocks perpendicular to the resolving direction, does not build per-block projections, and does not establish/interpolate pattern or lane boundaries. Crucially, the "interpolation" here is a regression within each lane's own band series, a materially different operation from connecting boundary points across blocks.
• "By another" nuance for § 102(e): This reference names Hara alone, whereas the subject patent names Hara and Sato — a different inventive entity, so § 102(e) status is arguable but generally available. Even so, because both are commonly owned by Fuji Photo Film, the reference is a candidate for disqualification as obviousness art under pre‑AIA § 103(c) — an important point if the art is used in a § 103 combination rather than as § 102 reference.

2.7 U.S. Patent No. 4,748,326

"Radiation image read-out apparatus" — Mori et al.; Fuji Photo Film Co., Ltd.
· Priority Dec. 27, 1985 · Issued May 31, 1988 · § 102(e)

• Description: Read-out apparatus for radiation image storage panels (scanning/optics/electronics). (Note: its Dec. 27, 1985 priority is exactly one year before the subject priority date of Dec. 27, 1986 — a coincidence, not a relationship.)
• Claim relevance: Apparatus background for the phosphor read-out clause of claims 1/12.
• § 102 verdict: No anticipation of any claim.
• Caveat: claim text not retrieved.

2.8 U.S. Patent No. 4,777,597 — the closest sibling signal-processing reference

"Signal processing method in autoradiography" — Shiraishi et al.; Fuji Photo Film Co., Ltd.
· Priority Jan. 8, 1983 · Issued Oct. 11, 1988 · § 102(e)

• Description (family-level retrieval): A method for obtaining one-dimensional locational information as symbols/numerals from digital signals corresponding to an autoradiograph recorded on a stimulable phosphor sheet (a film-based counterpart exists in the same Fuji family, e.g., the EP 0 113 675 / 0 113 677 group and JPS 59‑126530). It (i) forms a graph with scanning position on the abscissa and signal level on the ordinate, and (ii) applies smoothing and/or threshold processing to detect sampling points; a DNA base-sequence determination variant adds statistical processing of candidate sampling points.
• Claim relevance: Overlaps the preamble of claims 1/12 (phosphor-sheet digital signal → numerals/symbols) and the thresholding concept of dependent claims 4, 10, 15, 21 ("mean signal level … threshold value … point corresponding to the threshold value").
• § 102 verdict: No anticipation of claims 1, 7, 12, or 18. The reference detects sampling points (bands) on a scanning line; it does not divide the data into blocks perpendicular to the resolving direction, does not form per-block projections, and does not interpolate pattern/row boundaries between blocks. It could be used against the threshold-dependent dependent claims only in a § 103 combination, and even then only together with art supplying the block/interpolation features.
• Caveat: The exact US 4,777,597 claim text was not retrievable on this budget; the description above is inferred from the Fuji family to which the title and priority date attach and should be verified against the granted U.S. text before being relied on for a § 102/§ 103 position.

3. Summary table

4. Most relevant prior art, ranked

1. US 4,777,597 — nearest signal-processing sibling (graph of position vs. signal level, smoothing/threshold sampling-point detection, phosphor-sheet digital signals). Closest on the preamble and thresholding, but silent on block division/projection/interpolation.
2. US 4,665,312 — nearest on the "join detected points into a continuous line" concept (and on the traverse-then-scan idea), but its line is the resolving-direction scanline, not inter-block pattern/lane boundaries.
3. US 4,720,786 — same inventor as the subject patent; deals with plural resolved rows/lanes and lane-position variation, but corrects band positions along the resolving direction by intra-lane regression, not by cross-block boundary interpolation.
4. The remaining five (US 4,617,468; US 4,629,891; US 4,734,581; US 4,748,326; US 4,706,192) are infrastructure/background references (recording medium, read-out hardware/process, manual base-code entry) that support the preambles or general field but contain no boundary-detection or block-projection teaching.

5. Conclusion on § 102

No cited reference anticipates any of claims 1–23 of US 4,939,667. For every independent claim (1, 7, 12, 18), the cited art fails to disclose at least the combination of (a) dividing the image data into two or more blocks along a direction perpendicular to the resolving direction, (b) compiling per-block data into a perpendicular-position waveform (projection), and (d) interpolating the per-block boundaries into a total pattern boundary (plus the resolved-row boundary-and-interpolation steps of claims 12/18). Because each dependent claim incorporates all limitations of its independent claim, no dependent claim is anticipated either.

The prior art is therefore best deployed as:

• § 102(e)/§ 102(b) background for the environment (phosphor sheet, film, read-out, autoradiographic process, DNA base-sequence context) — refs. 1, 2, 4, 5, 7 and the preambles of 3, 6, 8;
• § 103 obviousness art, most plausibly by combining US 4,777,597 and/or US 4,665,312 (signal processing / point-to-line fitting) with US 4,720,786 (lane/row handling) — while remembering the common-ownership (§ 103(c)) overlay that applies to the Fuji-owned references, including the same-inventor US 4,720,786.

6. Caveats and limitations of this analysis

• Full claim texts retrieved for US 4,665,312, US 4,706,192, US 4,720,786 (and descriptions for US 4,777,597 at family level). Full claim texts were NOT retrieved in this session for US 4,617,468, US 4,629,891, US 4,734,581, US 4,748,326, or for US 4,777,597 itself; the descriptions for those are based on their titles/fields plus the subject patent's own citation entries (which are authoritative for the bibliographic data). Verify against the granted U.S. texts before relying on any § 102 position.
• All eight are U.S. patents; the patent's table lists no foreign patent documents among the "Patent Citations." (Foreign/EP family members such as EP 0 113 675/677 appear only through the sibling patents and are not separately listed as cited art.)
• The specification also references the applicant's co-pending U.S. applications Ser. Nos. 568,877 (abandoned), 664,405 (abandoned), 730,034, 837,037 (abandoned), 91, and 917,609 as related art/infrastructure. If any of these matured into U.S. patents, they are candidates for § 102(e) treatment (same-applicant, commonly-owned) and should be checked as a separate step; they are not part of the eight-item "Patent Citations" list analyzed here.
• The "Cited By" documents (e.g., US 5,972,619; US 6,208,941; US 6,241,080; US 11,133,087; WO 92/05427) post-date the patent and are not prior art; they are relevant only as evidence of the field's later development and were deliberately excluded from this § 102 analysis.
• This report addresses prior art only and does not revisit the litigation findings in the previously generated section; nothing here contradicts that "no litigation known" conclusion.

I'll ground this analysis in the prior-art list from the patent page. Let me pull details on the key cited references so I'm not characterizing them from titles alone.

Let me pull the two closest sibling references (US4894786 and US4837687) plus the "signal detecting" reference cited on the page.

One more targeted check on the closest sibling reference, US4894786, to pin down its date and disclosure.

Obviousness Analysis — U.S. Patent No. 4,939,667

Statute: 35 U.S.C. § 103, pre‑AIA version. The application (07/138,395) was filed December 28, 1987 and granted July 3, 1990; the AIA § 103 first‑to‑file regime applies only to applications filed on or after March 16, 2013. The Graham v. John Deere framework (scope/content of prior art; differences; PHOSITA level; secondary considerations) governs, as supplemented by KSR Int'l v. Teleflex (2007).

Ground rule carried over from the earlier sections: I do not repeat the litigation findings. Note also the cross‑reference caveat below.

0. Scope, record, and evidentiary limits (read this first)

This analysis is built on the "Citations (8)" list from the subject patent's Google Patents page (https://patents.google.com/patent/[US4939667](/patent/US4939667)/en) plus the claim text of US 4,939,667 itself (23 claims) and the description.

Honest limitation: I retrieved full or partial claim text for US4720786A and US4777597A (and European counterparts of the same Fuji January‑1983 priority group). For the remaining six cited references I have only bibliographic data (number, priority date, grant date, assignee, title) from the page. Where I characterize those six, I say so and mark it [title/bib only]. I have not verified the full disclosure of any reference against its specification, so the conclusions below are preliminary, not a formal invalidity opinion.

Items I use that are NOT among the examiner's eight citations are flagged as [adjacent document] — these come from the same Google Patents page (Similar Documents / family / cite‑by‑cite graphs) or from my searches, and are offered only as corroboration of what the art already knew.

1. Prior art from the page's "Citations (8)" list

Adjacent documents I rely on for corroboration only: EP 0 113 675 A2/A3 and EP 0 113 677 A2/A3 (Fuji, Jan. 1983 priority group); EP 0 123 943 B1; US 4,857,050A; US 4,837,687A; US 4,894,786A; and the JP family member JPS63167290A / JPH0570792B2 listed on the page.

One contradiction to flag: the page's "Similar Documents" table lists US4894786A, "Signal processing method for analyzing autoradiograph" (publ. 1990‑01‑16) — a different patent from a different Fuji filing, with a title nearly identical to the subject patent. It is not in the subject patent's family (family = US 07/138,395 + JPS63167290A + JPH0570792B2). Any analyst reusing this page should not conflate them. I have no filing date for it from the page, and therefore cannot date it for § 102 purposes; I exclude it from the combination.

2. The claimed subject matter, grouped

The 23 claims form four independent‑claim families that differ only in (i) data‑acquisition route and (ii) whether "resolved rows" (lanes) are separately bounded:

The only structural differences among the four independent claims are the acquisition route and the presence/absence of the lane‑boundary steps. Everything else — (a) split into ≥2 blocks along the axis perpendicular to the resolving direction; (b) per‑block compilation along the resolving direction to produce a position‑vs‑level waveform; (c) boundary from that waveform; (d)/(e) interpolation — is identical.

That symmetry matters for § 103: a single combination that renders claim 1 obvious also renders claim 7 obvious (by the known phosphor‑sheet ↔ film substitution), and the lane‑boundary pair (12/18) rises or falls together.

3. Element‑by‑element mapping to the cited art

Claim 1

Claim 12

Claim 12 adds "(d) using the waveform to establish a boundary of each resolved row in each block" and interpolates pattern‑area and row boundaries together. US4720786A is squarely on point: its FIG. 3 and claims 2/10 extract the digital signal along the resolving direction for each resolved row and locate band features by signal level; the same operation applied across an x‑projection yields the inter‑lane minima. Claim 16/22's "point at which the signal level is made minimum… within the pattern area" is the complement of US4720786's maxima.

4. The obviousness combinations

Combination 1 (primary): US 4,777,597 A in view of US 4,720,786 A

Yields: claims 1–6, 7–11, 12–17, 18–23 (all four families), subject to the film‑route substitution addressed in Combination 3.

Why a PHOSITA would combine them. Both references are directed to the identical problem — automatically converting a digitized electrophoresis autoradiograph into base‑sequence locational information. US4777597 supplies the pipeline and the diagnostic primitive (a position‑vs‑signal‑level waveform). US4720786 supplies the reason the pipeline fails on real gels: "smiling" and offset distortion make the resolved rows non‑parallel and non‑uniform, so that "the band spaces are dense in the upper region" and "the positions of bands are overall deviated from those of other rows." The subject patent's own specification concedes precisely this defect and the desired remedy: the image data "contains not only digital image data of information on a resolved pattern but also digital data of extra area… it is desired to first detect image data corresponding to a desired image area from the obtained digital image data and then perform the signal processing, so that quality and efficiency of the processing is enhanced" (US 4,939,667, col. 1, "Description of Prior Art"). An applicant's admission of the problem is strong evidence of the PHOSITA's motivation (In re Rosenberger; Graham factor 1–2).

Why the block‑wise projection + interpolation follows. Once the artisan accepts US4720786's teaching that distortion "varies locally in the direction of resolution," a single whole‑image projection is self‑defeating: summing two thousand rows whose lane walls drift laterally smears the inter‑lane minima into a featureless valley, whereas the specification itself notes that "the accumulated one‑dimensional waveform (projection) clearly contains sectional views of all 8 lanes" only because the drift is bounded within a block. Dividing into blocks and re‑projecting is the simplest known way to keep the projection valid over a locally quasi‑straight region, and then re‑joining the per‑block points by a line/curve is exactly what US4720786 already does per lane (regression line/curve, claims 3–4) and what EP 0 123 943 does with "straight lines, polygonal lines or curved lines."

KSR rationales engaged:

• (C) Known technique, improvement in the same way. US4720786's local‑correction technique is applied to the newly recognized task of isolating the pattern area — "use of [a] known technique to improve similar devices in the same way."
• (F) Obvious to try. A finite, small number of identified, predictable partitioning strategies existed; the spec itself says the number is "preferably 5 to 6," and claims only "two or more." That is a design choice with no demonstrated criticality.
• (A) Predictable result. Block‑wise averaging to sharpen a projection is standard signal processing; the output (sharper lane minima) is exactly what the artisan would expect.

Combination 2: Combination 1 + the read‑out/phos phor hardware references (US 4,735,581 / 4,629,891 / 4,748,326 / 4,617,468)

Yields: the preamble of claims 1 and 12 (and the corresponding elements of 7/18).

These references [title/bib only] supply the sheet, the read‑out method and the read‑out apparatus by which the digitized stimulated‑emission signal is generated. Because all four are Fuji Photo Film's dedicated disclosures of that hardware and were in the same program as US4777597, there can be no genuine teaching‑away; the combination is driven by the express purpose of the phosphor program itself — as the subject specification states, the autoradiograph "is stored in the phosphor sheet as radiation energy and then read out as stimulated emission in time sequence," so that "information can be expressed by the form of numerals and/or symbols." The whole point of the hardware was to feed the signal‑processing step. KSR rationale: (B) substitution/combination of known elements with predictable results.

Combination 3: Substitution of the radiographic‑film route (claims 7–11, 18–23)

Yields: the only structural distinction between the A/C families and the B/D families.

Both routes are expressly disclosed side‑by‑side in a single cited reference. US4720786A:

• claim 7: "digital signals… obtained by placing the support medium and a stimulable phosphor sheet… irradiating said phosphor sheet with stimulating rays and photoelectrically detecting the autoradiograph as stimulated emission"; and
• claim 8: "digital signals… obtained by placing the support medium and a radiosensitive material together in layers to record the autoradiograph… as a visible image and photoelectrically reading out the autoradiograph visualized on said radiosensitive material."

US4777597A teaches the same equivalence. Under KSR, "[w]hen a patent claims a structure already known in the prior art that is altered by the mere substitution of one element for another known in the field, the claim must do more than yield a predictable result." Here the substitution is not even an alteration — it is an alternative already recited in the same reference. The film‑route independent claims are the weakest claims in the patent.

Combination 4: The threshold limitations (claims 4, 10, 15, 21)

"a mean signal level of the waveform is determined, a threshold value is determined on the basis of the mean signal level, and a point corresponding to the threshold value on the waveform is assigned to the boundary."

The Fuji family's threshold‑based sampling‑point detection — "processing said graph with either or both of smoothing and threshold‑processing to detect sampling points" (EP 0 113 677 [adjacent]) — discloses thresholding a position‑vs‑level graph to establish a feature coordinate. Using the mean of the waveform as the threshold basis is a routine statistical design choice with a predictable result (a threshold that adapts to overall signal level), and it is described in the subject specification as merely exemplary: "obtaining a threshold value (z_a) by multiply the average an appropriate number"—the multiplier itself being left to the operator. KSR: obvious design choice / result‑effective variable absent. The claim recites no critical multiplier value and no unexpected result.

Combination 5: The interpolation‑geometry limitations (claims 5, 11, 17, 23)

"the boundary… is determined on a central line… for each block and boundary points of the blocks which correspond to each other are connected in sequence."

Grounding: EP 0 123 943 B1 [adjacent] — "joining the corresponding reference sampling points among said plural reference rows to fix plural continuous lines comprising straight lines, polygonal lines or curved lines" — and US4720786A's per‑lane regression‑line/curve fitting (claims 3–4). Placing the sample point on the block center line and connecting like‑numbered points across blocks is the textbook piecewise‑linear reconstruction of a smooth boundary from a small number of samples. No criticality is asserted, and the specification gives no data on spacing or block count. Obvious.

Combination 6: The lane‑minimum limitation (claims 16, 22)

"a point at which the signal level is made minimum of the waveform in the pattern area."
US4720786A claims 2 and 10 locate maxima ("positions where level of the extracted signal is maximum") along the resolving direction of each row. Detecting the complementary extremum — the valley between adjacent lanes of a projection — is the same operation with the opposite sign, and the specification itself defines the operation generically: "The minimum point can be determined by detecting a point on which the sign of difference inverses." Obvious.

Combination 7: The biopolymer/base‑sequence limitation (claim 6)

Expressly disclosed: US4777597A, claims 8–9 — "biopolymers are nucleic acids, derivatives of said nucleic acids or cleavage products of said nucleic acids; and said symbols, numerals and combinations… represents a base sequence of biopolymers"; and US4720786A, claim 6 — the four groups (guanine‑, adenine‑, thymine‑, cytosine‑specific DNA fragments). Anticipated/obvious.

Combination 8: The compilation‑mode limitations (claims 2, 8, 13, 19 and 3, 9, 14, 20)

• Compile all data along the resolving direction (2/8/13/19): the plain meaning of "projection"; US4720786A's FIG. 3 waveforms; and the subject specification's own description of adding "all of the image data contained in the (a) block."
• Extract at equal spaces (3/9/14/20): the specification itself justifies it purely as a labor‑saving optimization — "signals corresponding to a portion of the pixels can be added for simplifying the addition operation… only 10 to 50% of the whole image data can be subjected to the addition operation." Under KSR, a claim whose stated benefit is reduced computation, with no unexpected result, is obvious; uniform sub‑sampling was ubiquitous in 1980s image processing.

Combination 9 (secondary): US 4,706,192 A (Hitachi Software Engineering)

An "input apparatus for entering base sequence information of the gene" [title/bib only] supplies the digital base‑sequence data‑entry/editorial environment and evidences that machine‑handling of sequence information with operator input was known in the 1984–87 window. It is a plausible secondary reference for the "in the form of numerals and/or symbols" purpose and for claims‑drafting context, but it is weaker than Combinations 1–3 and I would not rest a rejection on it alone.

5. Graham factor 4 — secondary considerations

I found no evidence of objective indicia in the record on this page:

• No unexpected results. The asserted advantage — accurate extraction "even when the resolved pattern and resolved rows are curved or inclined to one side" — is the predicted consequence of locally valid projections, and the specification reports no comparative data against a non‑blocked method.
• No copying / licensing. (Consistent with the earlier no‑known‑litigation finding — I do not restate it here.)
• Commercial success, if any, is not nexus‑qualified. Fuji's imaging‑plate (BAS) systems and the underlying phosphor technology were driven by US4777597A, US4734581A, US4629891A, US4748326A and US4617468A — the acquisition inventions — not by the block‑and‑interpolate boundary‑detection algorithm of the claims here. The claims cover a data‑processing step buried inside a commercial instrument; any success is attributable to the instrument and the phosphor chemistry.
• Long‑felt need / failure of others: not evident. US4720786A shows the same team was actively solving the adjacent distortion problem in the same window; there is no teaching‑away.

6. Anticipated rebuttals and my assessment

7. Bottom line

1. Prima facie, all 23 claims are obvious under pre‑AIA § 103 over the combination of US 4,777,597 A + US 4,720,786 A, further in view of the Fuji read‑out/phosphor‑sheet references (US 4,735,581 / 4,629,891 / 4,748,326 / 4,617,468) for the acquisition preamble and, if desired, US 4,706,192 for the sequence‑data handling environment.
2. The film‑route claims (7–11, 18–23) are the most exposed, because US4720786A itself recites both the phosphor‑sheet and radiographic‑film data routes in adjacent claims (claims 7 and 8), making the substitution a mere alternative explicitly disclosed in a single reference.
3. Treating the threshold, mean‑signal‑level, minimum‑signal‑level, equal‑space sampling and "central line" limitations as merely exemplary is well supported: the specification describes each as an optimization without criticality, and older claims carry no data.
4. The single best candidate for survival is the specific block‑division‑plus‑interpolation architecture, particularly as applied to per‑lane (resolved‑row) boundaries in claims 12–17 and 18‑23 — but that survival depends entirely on whether US4720786A's localized‑distortion teaching is accepted as supplying the suggestion to partition along the perpendicular axis. I assess this as a close question, slightly favoring the challenger, given the applicant's own admission in the specification that the pattern and lanes may be curved or inclined and that extracting the desired image area was already a recognized desideratum.

Confidence: Moderate for the overall § 103 conclusion; high for the mapping of the acquisition step and the film‑substitution theory (both textually grounded); low‑to‑moderate for the six references I could only characterize from titles and bibliographic data, which I flagged [title/bib only] throughout. A definitive opinion requires the full text of US 4,665,312, US 4,735,581, US 4,629,891, US 4,748,326, US 4,617,468 and US 4,706,192 — none of which I verified in this session — and confirmation of the EP 0 123 943 and EP 0 113 675/677 publication dates against the EPO register.