US 4972325

USPTO / CAFC Lookup and Patent Summary — US 4,972,325

Part 1 — Search results for the specific number 4972325

Negative on both targets, with a sourcing caveat I am holding to.

USPTO database. I could not open a live Patent Public Search / Patent Center / Assignment Center query in this session (step limit reached before the direct interface queries completed). What I did retrieve, and what is authoritative for the bibliographic facts below, is the full text of the patent record served from USPTO-derived mirrors: patents.google.com/patent/US4972325 and uspto.report/patent/grant/4972325. Both return a single, consistent record — title, inventor, assignee, application 07/031,841, filing 1987‑03‑30, issue 1990‑11‑20 — with no 2026 prosecution, post‑grant, or reissue activity.

CAFC 2026 dockets. No 2026 Federal Circuit matter keyed to 4972325 was returned. Every literal 4972325 hit in the search results is an unrelated string, not a patent docket, and per the strict no-auto-correction rule I am expressly not treating any of them as this patent:

Conclusion: no USPTO 2026 proceeding and no CAFC 2026 appeal involving US 4,972,325 were found. This is consistent with the earlier-generated litigation and PTAB sections (both of which found zero contested activity). The absence is expected: the patent is recorded "Expired – Lifetime," anticipated expiration 2007‑11‑20, eight years before the task's stated current date.

Uncertainty I will not paper over: my negative rests on the patent full text plus open web sources, not on a completed direct docket-database query. For belt-and-suspenders confirmation, run USPTO Patent Public Search (https://ppubs.uspto.gov/), Patent Center (application 07/031,841), and the CAFC docket / PACER Case Locator by the literal number.

Part 2 — Concise patent summary

Abstract (verbatim, then translated)

"A signal processing method for determining base sequence of nucleic acids by subjecting digital signals to signal processing, said digital signals corresponding to an autoradiograph of plural resolved rows which are formed by resolving a mixture of base-specific DNA fragments or base-specific RNA fragments labeled with a radioactive element in one-dimensional direction on a support medium, which comprises steps of: (1) preparing one-dimensional waveform composed of position along the resolving direction and signal level for each resolved row; (2) dividing the one-dimensional waveform into at least two intervals; and (3) smoothing the one-dimensional waveform in every interval through a smoothing means having different characteristics."

Plain language: take the scanned autoradiograph of a DNA-sequencing gel; trace a signal-level profile down each lane; cut each lane's profile into segments; and smooth the noise out of each segment by a different amount — more smoothing where bands are far apart (bottom of the gel), less where they are crowded together (top of the gel). Then detect and order the bands to read off the sequence.

Independent claims in plain language

Claim 1 — the "variable-mask moving average" claim. Read a four-lane (or more) sequencing autoradiograph; for each lane build a 1‑D waveform of position along the migration direction vs. signal level; pick at least two intervals along that waveform; run a moving average filter whose mask size differs from interval to interval; then run the smoothed signals through further processing to detect and sequence the bands (i.e., output the base order). The hallmark is the moving-average mask size that changes across intervals — and per claim 3, the mask grows as migration distance grows.

Claim 7 — the broader "smooth according to band spacing" claim. Same front end (lane waveforms, at least two intervals selected along the resolving direction), but step (3) recites a generic smoothing means that smooths each interval "according to spaces of bands in each said interval" — it does not require a moving average or a "mask size." It then requires explicitly detecting the band sequence in each lane and determining the base sequence from it. Because the filter type is generic, claim 7 is the broader of the two independent claims. Its dependent claims specify the alternative smoothing means as a weighted mean filter (claim 8) or a low-pass-characteristic filter (claim 9).

Dependent-claim map (for orientation)

• On claim 1: claim 2 (mask varied in every interval), claim 3 (mask enlarged with resolving distance), claim 4 (the four G/A/T/C fragment groups → four lanes), claim 5 (data obtained via stimulable phosphor sheet, laser-scanned as stimulated emission), claim 6 (data obtained via radiosensitive material / film, photoelectrically read out).
• On claim 7: claim 8 (weighted mean filter), claim 9 (low-pass filter), claim 10 (G/A/T/C four-group scheme), claim 11 (stimulable phosphor read-out), claim 12 (radiosensitive material read-out — note the record spells it "radiosenstive," left uncorrected).

Caveats and cross-references

• "EP 0199327 / EP 0242629 / US 4,980,827" trap. The specification's own "Similar Documents" neighbourhood contains references that look relevant; as the earlier prior-art and obviousness sections flagged, US 4,980,827 is not available prior art against this patent despite being the most on-point document in the field. I do not repeat that analysis here.
• Assignee rendering. I use "Fuji Photo Film Co., Ltd." as the operative name and quote the literal "FUJI PHTO FILM" record only when quoting it — consistent with the earlier assignment section.
• Date discrepancy, flagged again. The header of this task states April 26, 2026; the authoritative record I fetched carries a retrieval stamp of 2026‑09‑26T08:18Z, and the earlier PTAB section used 2026‑09‑27. Nothing in this summary depends on which date applies; the patent expired in 2007, so all three dates are uniformly after term.
• Confidence. Title, inventor, assignee, filing/issue dates, abstract, and the claim language are taken verbatim from the authoritative full text and are high-confidence. The USPTO/CAFC "no 2026 activity" result is high-confidence in substance but rests on open sources rather than a completed first-party docket query, as noted in Part 1.

I'll search for litigation involving US patent 4972325 across patent litigation databases.

Let me search the specialized patent litigation databases directly.

Search Results: Litigation Involving U.S. Patent No. 4,972,325

Bottom line: I found no record of any litigation (district court, ITC, or CAFC) involving U.S. Patent No. 4,972,325.

What I searched

I ran multiple query variations against patent-litigation and general web sources, including the patent number in both formats (with and without commas):

• "4972325" and "4,972,325" combined with litigation/infringement terms
• portal.unifiedpatents.com litigation "4972325" (Unified Patents litigation caselist)
• docketalarm "4,972,325" Fuji Photo Film
• Fuji Photo Film patent 4972325 lawsuit DNA sequencing
• Direct checks of the Google Patents, Justia Patents, FreePatentsOnline, and uspto.report pages for the patent.

What those sources showed

• No litigation section exists on any of the patent record pages. The Google Patents entry for US4972325A lists only prosecution/assignment history and citation relationships — no "Litigation" or "Related Litigation" entries.
• The "Cited By" and "Citations" tables are patent-to-patent citations, not litigation. These include US5119316 (DuPont), US5365455 / US5502773 (Vanderbilt), US5585720 (Western Atlas), the Hyseq/Affymetrix family, WO1993005183, EP0592060, etc. None of these are lawsuits against the '325 patent; they are later patents that cite it as prior art.
• The "Similar Documents" list (e.g., US4802101, US4941092, US4958281, US4980827, EP0199327B1, EP0242629B1, JP publications) reflects family/sibling and same-field Fuji Photo Film patents, not adversarial proceedings.
• The 4972325 strings that do surface in search results are unrelated to the patent: an SEC 8-K accession identifier (Eloxx Pharmaceuticals), a Florida LLC FEIN (99-4972325), a Brazilian trademark application number, a geodetic coordinate value, and a UK probate notice number. The strict "do not auto-correct / do not return similar numbers" rule means none of these should be confused with U.S. Patent 4,972,325.

Required fields (per your request format)

Because no case was identified, there is nothing to populate:

Contextual notes bearing on the "no litigation" conclusion

1. Patent term. U.S. 4,972,325 issued November 20, 1990, from Application No. 07/031,841 filed March 30, 1987, claiming priority to JP 61-71881 (March 29, 1986). The Google Patents record shows "Anticipated expiration 2007-11-20" and status "Expired – Lifetime." Any infringement action would have had to be filed (and could only concern conduct occurring) within that window.
2. Assignee track record. The patent is assigned to Fuji Photo Film Co., Ltd. (later Fujifilm Holdings Corp.). Fuji Photo Film is a well-documented patent litigant in other matters (e.g., Fuji Photo Film Co. v. International Trade Commission, Fed. Cir. 2004/2007, cited in my search results), but that ITC/CAFC case concerns other Fuji patents, not the '325 patent.
3. Subject-matter reality. The '325 patent claims a signal-processing method for reading DNA-sequencing autoradiographs (interval-divided, variable-mask moving-average smoothing). This is a method claim in a niche, laboratory-equipment context — a low-litigation profile.

Confidence and caveats

• Confidence that no litigation exists is moderate-to-high but not absolute. My queries did not return results from a paywalled or fully indexed PACER/CourtListener/Unified Patents page keyed to this exact number. If litigation exists but was never indexed in the free sources I can reach, I would not have seen it.
• Recommended direct verification: Search PACER (Case Locator) and CourtListener/RECAP by the exact patent number 4,972,325, and check the Unified Patents litigation portal and Docket Alarm by patent number, not by party. I was unable to complete the final round of targeted docket-database queries within this session, so those two sources are the highest-value places to confirm.
• No results should be fabricated. I am reporting the absence of records in the sources I actually retrieved, not asserting a definitive negative.

If you want, I can attempt a further round of targeted docket-database searches (CourtListener RECAP, Unified Patents, Docket Alarm) with alternate query formatting to try to confirm the negative result.

I'll verify the canonical "no PTAB proceedings" default with targeted searches before finalizing.

Let me run one more targeted confirmation round, since the number may appear in PTAB filings in different formats.

PTAB / AIA Trial Proceedings — US 4,972,325

Verification note before the findings. I sourced the proceedings list from the structured "PTAB proceedings on file" block (USPTO Open Data Portal ingest), which returns no AIA trial proceedings. I then ran independent web searches on the literal identifier in multiple formats (4,972,325, US4972325, "4972325" + PTAB / IPR / petition) to catch anything the ODP has not indexed. Every hit keyed to that number resolved to either (a) the patent's own Google Patents / FreePatentsOnline / uspto.report page, or (b) an unrelated 4972325 string — a Florida LLC FEIN, an Eloxx Pharmaceuticals SEC 8-K accession, a Brazilian trademark filing, a Coal India tender number, and a Russian geodetic coordinate. None is a PTAB proceeding. I flag explicitly that I did not open a PTAB E2E (PTACTS) docket view keyed to this patent in this session; the negative rests on the structured block plus web search.

Proceedings overview

Total count: 0 AIA trial proceedings on file. Breakdown: 0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled, 0 institution-denied — because no IPR, PGR, or CBM petition was ever filed against US 4,972,325, and no ex parte reexamination or reissue is reflected in the post-grant legal-events record either.

Bottom line for a defendant: this is not "the patent survived two IPRs and is hardened," and it is not "claims 1–5 were canceled." The correct reading is that the entire claim set (claims 1–12) is UNTESTED at the PTAB, and untested because there is no economic reason left to test it. The patent's 20-year term ran out on 2007-11-20 (filing 1987-03-30), roughly five years before IPR and PGR became available on 2012-09-16. A demand letter citing US 4,972,325 in 2026 therefore has no live enforcement theory — and correspondingly, a defendant's incentive to file an IPR (which exists to knock out a claim you are being sued on) evaporates.

No proceedings — detail

There is no proceeding number to report. Under the constraints in your instructions, I will not synthesize one. The correct entry is a null:

Why the absence is structural, not accidental. Four independent bars explain a zero-complaint PTAB history, and they compound:

1. Term expired before the forum existed. IPR and PGR became available 2012-09-16; CBM likewise, with CBM sunsetting for new petitions on 2020-09-15 under AIA § 18 and 37 C.F.R. § 42.300(d). US 4,972,325 was already Expired – Lifetime (anticipated expiration 2007-11-20) before any of those windows opened. The Board can, in principle, entertain a petition against an expired patent, but the patent owner cannot amend claims, and — critically — there is no infringement exposure to defeat.
2. PGR was categorically unavailable. PGR reaches only patents having a claim with an effective filing date on or after 2013-03-16. This patent's priority date is 1986-03-29 (JP 61-71881); filing date 1987-03-30. PGR jurisdiction never attached.
3. CBM standing could never be satisfied. Even during the 2012–2020 CBM window, a petitioner had to have been sued for infringement or charged with infringement of the patent (AIA § 18(a)(1)(B)). With the patent expired since 2007, that predicate could not be met for live conduct.
4. IPR § 315(b) time-bar was moot, and so was the motivation. The § 315(b) one-year clock runs from service of a complaint alleging infringement. No such complaint is on file (consistent with the litigation section: no district court, ITC, or CAFC activity), so there was never a trigger — or a reason.

Contrast worth noting (not this patent). Fuji Photo Film / Fujifilm, the assignee here, is a repeat PTAB litigant in other matters — e.g., IPR2017-01356, Fujifilm Corp. v. Sony Corp., on U.S. Patent 7,016,137 (oral hearing requested 2018-08-31; Docket Alarm). That proceeding concerns an unrelated tape/recording patent. I raise it only to show that the assignee's zero-IPR history on this patent is attributable to this patent's age and subject matter, not to any institutional reluctance by the family to appear at the Board. Likewise the ITC/CAFC proceedings around U.S. Patent 4,972,649 (Fuji Photo Film Co. v. Int'l Trade Comm'n, 386 F.3d 1095 (Fed. Cir. 2004)) involve a different patent (lens-fitted film packages) and are not proceedings on US 4,972,325 — the similar-looking number is a trap; per your operating rules I have not auto-corrected it into an identity.

Strategic summary

Claim-status picture. All twelve claims — independent claim 1 (moving-average filter with interval-varied mask size) and independent claim 7 (any smoothing means matched to band spacing), plus dependents 2–6 and 8–12 — are UNTESTED. None is CANCELED; none is SUSTAINED by any administrative or judicial body. If someone tells you claims of the '325 patent have been invalidated, ask for the FWD — there isn't one. The only patentability determination this claim set has ever received is the original 1990 ex parte allowance (grant date 1990-11-20).

Estoppel landscape: empty, but useless. Because no IPR was instituted, no § 315(e)(2) estoppel ever attached. Practically, a defendant today faces no estoppel and may raise any § 102/§ 103/§ 112 ground, in any forum, on any art — including the five references cited on the patent's own face (US 4,777,597; US 4,835,733; US 4,665,312; US 4,720,786; US 4,802,101), all Fuji Photo Film, and all now long expired as well. But this freedom is theoretical: with the '325 term over since 2007-11-20, there is no injunctive relief and no recoverable damages. Under 35 U.S.C. § 286, recovery reaches back only six years from filing, i.e. conduct after roughly 2020 — every day of which post-dates expiration.

Pattern signals. Zero repeat petitions (the classic "same petitioner, multiple IPRs" pattern is absent); zero PTAB appeals by the patent owner on this patent; no defensive aggregator (Unified Patents or similar) in the chain for this number. The assignee chain — original Fuji Photo Film Co., Ltd., current database attribution Fujifilm Holdings Corp — has no recorded reassignment event for this patent beyond the original 1987 inventor assignment (Reel/Frame 004704/0707, effective 1987-03-27). This is a dormant, expired, single-assignee patent cited 20+ times as prior art by later sequencing-analytics patents (Affymetrix, Hyseq/Callida, DuPont, Vanderbilt, Baylor) — citation is not litigation and is not PTAB activity.

One contradiction to flag. My earlier section noted that the working prompt's stated "current date" (April 26, 2026) differs from a retrieved result date (2026-09-27). Nothing in the PTAB analysis turns on it; I use 2026-09-27 here per the task header.

Recommended next steps

If you are a defendant receiving an assertion of US 4,972,325:

• Do not spend money on an IPR. The patent expired 2007-11-20. There is no viable § 271 theory for conduct after that date, and § 286 forecloses reaching back to the term. The correct first response is a term-expiration / no-recoverable-damages position, not a post-grant petition. Quoting the record: Google Patents' legal-status entry reads verbatim "Expired - Lifetime" with "2007-11-20 — Anticipated expiration." There is no FWD to link to and no disposition to quote, because no proceeding exists.
• Preserve the null as evidence. The structured ODP block returning zero AIA trials, coupled with the absence of any district court, ITC, or CAFC matter (per the litigation section), is itself the defensive narrative: this patent has never been asserted or challenged in a contested forum.
• If a demand letter affirms claims 1–12 as "valid and enforceable": that is technically true in the sense that no tribunal has held otherwise — but it is materially misleading if offered as a threat, since expired claims cannot be infringed prospectively. Ask the sender to identify the accused acts and their dates.

If you are monitoring rather than defending:

• Nothing to docket-watch. There are no trial-stage milestones (no institution decision deadline, no oral hearing, no statutory 1-year FWD due date) because no trial was instituted. There is no § 315(b) clock running on anyone.
• If you nonetheless want belt-and-suspenders confirmation, run the direct queries I could not complete in this session: PTAB E2E / PTACTS (https://ptacts.uspto.gov/) by application number 07/031,841 and by patent number; USPTO Patent Center for any reexamination or reissue certificate; and CourtListener / RECAP (https://www.courtlistener.com/) for any CAFC or district matter. I found no reexamination or reissue in the retrieved legal-events record — the only post-grant entries are maintenance-fee payments at years 4 (1994-05-20), 8 (1998-05-20), and 12 (2002-04-04) — but I did not exhaustively confirm a negative on reexam through Patent Center.

Plain statement of the negative: There is no PTAB activity on US 4,972,325. The absence is not a sign the patent is bulletproof; it is a sign the patent is old enough that the AIA trial regime post-dates its death.

I'll verify the assignment record against the USPTO Assignment Center and check for any post-issuance transfers.

Let me check the assignment record pages and any post-2006 change-of-name recordings.

Let me try to retrieve the recorded assignment's correspondent details and check for any Fujifilm re-recording.

Assignment & Ownership Analysis — US 4,972,325

Method and limitation up front. I attempted to pull the record directly from the USPTO Assignment Center (assignmentcenter.uspto.gov) and its predecessor index (assignment.uspto.gov/patent), plus uspto.report, FreePatentsOnline, and Justia. The Assignment Center search interface was not directly queryable in this session; what I can ground is (a) the assignment data printed in the patent's own front matter / Google Patents legal-events block (authoritative), and (b) corroborating third-party patent database pages. Accordingly, the "correspondent of record" field — the single most probative field for your NPE-attorney cross-check — is NOT recoverable from the sources I could reach, and I will not guess at it. Reel/frame and dates are reported as recorded.

Inventors

Unusual-pattern check — negative. No inventor-departure signal is present or even available: this is a single-inventor patent, and the inventor was a Fuji employee working in Fuji's radiation-image / autoradiography research program. Corroborating evidence that Hara remained at Fuji well past filing: he appears as inventor on later Fuji Photo Film applications, e.g. EP 1,122,665 (applicant Fuji Photo Film Co., Ltd.; inventor "Hara, Makoto, Ashigarakami-gun, Kanagawa-ken, JP"). A same-name inventor still prosecuting under the same corporate applicant years later is inconsistent with a post-filing talent drain. There is no 12-month all-inventors-depart portfolio-fire-sale pattern here — it is arithmetically impossible with one inventor, and the continuation of Fuji filings points the other way.

Original assignee

Fuji Photo Film Co., Ltd. — recorded in the USPTO assignment datablock literally as "FUJI PHTO FILM CO., LTD., A CORP. OF JAPAN" (the "PHTO" spelling is a transcription artifact in the record; I am not correcting it, per the literal-identifier rule. The same "FUJI PHTO FILM" rendering appears twice in the recorded text, once with the "A CORP. OF JAPAN" qualifier).

• Primary line of business: photographic film, photographic paper, photofinishing chemistry, and — material to this patent — radiation image recording and reproducing systems built on stimulable phosphor (BaFBr:Eu²⁺) sheets, i.e., the computed-radiography plate reader line. Claim 5 of this patent claims the method as practiced with exactly that Fuji hardware.
• Did they ship a product embodying the claims? Yes — highly likely, though I did not verify a specific model number in this session. The claims recite a signal-processing method, and the specification describes it as the pre-processing stage of an automated DNA-sequencing reader operating on a stimulable phosphor sheet read out by laser scanning. Fuji built and sold that class of instrument (the BAS / bio-imaging analyzer line arose from this program). The dependent claims (5, 11) are drafted around Fuji's own proprietary phosphor-sheet read-out, which is a strong indicator the claims were written to cover Fuji's own commercial apparatus rather than a hypothetical licensee's.
• Current status: Operating, restructured. Fuji Photo Film Co., Ltd. ceased to exist as the parent on 2006-10-01, when it converted to a pure holding company renamed Fujifilm Holdings Corporation (Fuji Photo Film → holding company; the photographic/imaging operating business transferred to a newly formed Fujifilm Corporation). Confirmed by ADR depositary notice (name/CUSIP change effective 2006-10-02) and contemporaneous trade-press coverage of the "Second Foundation" reorganization. Not acquired by a third party, not dissolved, not in bankruptcy.

Assignment timeline

One — and only one — assignment is recorded against this patent.

• 1987-03-27 (executed) / recorded 1987-03-30 — Reel 004704 / Frame 0707
  • Conveyance: Assignment of assignors' interest ("ASSIGNMENT OF ASSIGNORS INTEREST")
  • Assignor: HARA, MAKOTO (sole inventor)
  • Assignee: FUJI PHTO FILM CO., LTD., A CORP. OF JAPAN (literal spelling as recorded)
  • Correspondent: NOT RECOVERABLE from the sources available to me. The reel/frame is known (004704/0707) but the recorded correspondent-of-record attorney/firm is not exposed in the patent front matter or in any indexed third-party record I could reach. I decline to infer it. Note that the corresponding prosecution-side agent on comparable 1987-era Fuji US filings was the firm Sughrue, Mion, Zinn, Macpeak & Seas (seen in the record for other Fuji US registrations), but that is a prosecution attorney observation from an adjacent file — it is not the reel 004704/0707 correspondent and must not be treated as such.
  • Context: In-house assignment from employee-inventor to employer at filing — a routine pre-issuance employer-rights assignment. Filed the same week the application was lodged (application 07/031,841 filed 1987-03-30). Not an acquisition, not a fire sale, not a securitization.

Post-issuance: nothing. There is no recorded change of name, merger, security agreement, license, release, correction, or reassignment in the chain. The Google Patents legal-events block for US4972325A shows only: the 1987 assignment, the 1990 grant, four maintenance-fee payments (years 4/8/12 — 1994, 1998, 2002), and the 2007-11-20 anticipated expiration. Maintenance fees were paid in full through year 12, and the patent expired on its own term — it was never abandoned for failure to pay and never sold.

Handling note on the "Current Assignee" field. Google Patents displays the current assignee as "Fujifilm Holdings Corp." I want to be precise about what that is and is not: it is a database attribution inferred from the 2006 corporate reorganization, not a reassignment event I verified as recorded against this patent's file. The USPTO record as I can see it still names Fuji Photo Film as assignee of record. The material fact for your purposes is that either way the chain terminates inside the Fuji corporate family — it does not terminate at a third party.

Timeline diagram

timeline
    title Ownership of US 4972325
    1986 : JP priority application filed
    1987 : US application 07-031-841 filed
         : Inventor Hara assigns rights to Fuji Photo Film
    1990 : US patent issued
    1994 : Maintenance fee year 4 paid
    1998 : Maintenance fee year 8 paid
    2002 : Maintenance fee year 12 paid
    2006 : Fuji Photo Film becomes Fujifilm Holdings
    2007 : Patent term expires

NPE / troll-pattern signals

Verdict

Insufficient data — with a strong operating-company-retention reading.

Per your rubric, "Insufficient data" covers the case where the record shows only the original assignment, which is exactly what Reel 004704/0707 (executed 1987-03-27, recorded 1987-03-30) shows: a single inventor-to-employer conveyance, followed by nothing for the remaining 20 years of the term. Not one of the eight NPE signals is present, and the two that could theoretically be "unclear" (repeat correspondent, asserter identity) are structurally unavailable because there is no second link to compare against. The chain terminates at Fuji Photo Film Co., Ltd. / Fujifilm Holdings — an entity that manufactured the stimulable-phosphor autoradiography apparatus the claims were drafted to cover, and that paid all maintenance fees through year 12 rather than abandoning or monetizing the patent. This is the profile of a captive, defensive, in-house portfolio patent, not an NPE asset. The one meaningful gap in the analysis is the unrecovered correspondent on Reel 004704/0707, which is also the only place a subsequent NPE-lawyer footprint could have hidden had the record been incomplete.

Verification link: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search by patent number 4972325 or application 07/031,841); legacy index mirror — https://assignment.uspto.gov/patent/index.html. A direct query on either should surface Reel 004704/0707 and, critically, the correspondent field I was unable to retrieve — that field is the single item that would either close this file or reopen it.

Caveats I am holding to: (1) I did not fabricate or infer an assignment; the single link is the whole chain as recorded. (2) The "Fujifilm Holdings Corp" current-assignee attribution on Google Patents is a database-merge artifact of the 2006-10-01 reorganization, not a reel/frame I verified. (3) One earlier-generated section of this analysis notes a possible date-source inconsistency in the working set (April 26, 2026 vs. a 2026-09-27 retrieval stamp); nothing in this ownership analysis turns on it.

I'll research each cited prior-art reference directly. Let me run targeted searches on the patent and each of its five face citations.

Prior-Art Analysis for U.S. Patent No. 4,972,325

Scope note

The patent identifier searched is exactly 4972325 (U.S. Patent No. 4,972,325). I did not substitute any similar number. The authoritative full text supplied in my working set (Google Patents, https://patents.google.com/patent/[US4972325](/patent/US4972325)/en) is the source of truth, and it is consistent with the search results retrieved for the individual references. Two clarifications carried over from the earlier sections (not repeated here):

• "Citations" ≠ "Cited By." The five references examined below are the ones cited on the face of the '325 patent (prior art the examiner considered). The separate "Cited By" list (DuPont, Vanderbilt, Hyseq/Affymetrix, etc.) is later art that cites the '325 patent — it is not prior art to it and is not analyzed here.
• Date discrepancy: the task states April 26, 2026 while one system field reads 2026-09-27; this does not affect any finding below.

I was unable to open USPTO PatentCenter directly in this session (tool-step limit), so the bibliographic data below is taken from the patent full text and corroborating patent databases rather than a live USPTO query. I flag where a description is inferred rather than read directly.

The five cited references at a glance

Framework. Because the '325 application was filed March 30, 1987 and granted November 20, 1990, the pre-AIA version of 35 U.S.C. § 102 governs. All five references issued after the '325 filing date (March 30, 1987), so none is § 102(a)/(b) art on its face; each can only be prior art under § 102(e) (a U.S. patent granted on an application filed before the applicant's invention). Under In re Hilmer, the § 102(e) date is the actual U.S. filing date, not the foreign (JP) priority date.

Reference-by-reference analysis

1. US 4,777,597 — "Signal processing method in autoradiography"

• Full citation: U.S. Patent No. 4,777,597, inventors Hisashi Shiraishi et al., assignee Fuji Photo Film Co., Ltd.; application Ser. No. 07/024,909; JP priority 58-1328/83.
• Dates: JP priority 1983-01-08; granted 1988-10-11.
• Description: Digital-signal processing of an autoradiograph. The specification teaches preparing a one-dimensional graph of scanning position vs. signal level, and processing that graph by smoothing and/or threshold processing to detect sampling points, with the digital data obtained via a stimulable phosphor sheet. Applied to determining DNA base sequence. This is also discussed in the '325 specification's own background.
• § 102 status: Granted after the '325 filing → § 102(e) art (U.S. filing-date controlled).
• Potential § 102 relevance: Discloses the "one-dimensional waveform" step and the "further processing to detect bands" step, and teaches stimulable-phosphor read-out (the feature recited in '325 claims 5 and 11). It does not disclose (a) dividing the waveform into at least two intervals or (b) a moving-average filter whose mask size is varied per interval.
• Anticipation conclusion: Does not anticipate claim 1 or claim 7. Those independent claims require interval division and interval-varying smoothing strength, which this reference lacks (it uses a single smoothing/threshold pass). It is at most background/§ 103 art and could be relevant to the read-out-dependent features of claims 5/11 — but those claims are dependent and rise or fall with claim 1.

2. US 4,837,733 — "Signal processing method in autoradiography"

• Full citation: U.S. Patent No. 4,837,733, assignee Fuji Photo Film Co., Ltd.; continuation of Ser. No. 06/568,906 filed 1984-01-06; JP priority 58-1328/83.
• Dates: JP priority 1983-01-08; granted 1989-06-06.
• Description (directly retrieved): Determines a scanning line, then detects sampling points. The signal is "smoothed, for instance, by convolution of the signal graph using a suitable filter function" — the specification shows a single filter function g(W) in FIG. 7 and the smoothed result h(W)=g(W)*f(W) in FIG. 8, followed by thresholding and histogram processing. Also covers multi-row DNA base determination and stimulable-phosphor read-out.
• § 102 status: § 102(e) (US parent filed 1984-01-06).
• Potential § 102 relevance: This is the closest of the five on the smoothing element, because it expressly uses convolution/filtering of the one-dimensional signal to detect sampling points. However, it uses one fixed filter function applied to the entire signal — there is no division into intervals and no interval-varying mask size.
• Anticipation conclusion: Does not anticipate claims 1 or 7. The distinguishing limitation of the '325 patent (mask size / smoothing strength varied between intervals) is absent. This reference is best characterized as the primary § 103 starting point that the '325 applicant improved upon.

3. US 4,665,312 — "Signal detecting method in autoradiography"

• Full citation: U.S. Patent No. 4,665,312, inventors Shiraishi et al., assignee Fuji Photo Film Co., Ltd.; JP priority 57,415/83.
• Dates: JP priority 1983-03-31; granted 1987-05-12.
• Description: Directed to read-out efficiency — preliminarily scanning a stimulable phosphor sheet (low-intensity) at two or more positions to determine the one-dimensional distribution direction of the labeled substances, then scanning along that direction to obtain the locational information as a digital signal.
• § 102 status: § 102(e) (US filing before applicant's invention; grant 1987-05-12 is just after the '325 filing of 1987-03-30).
• Potential § 102 relevance: Relevant to how the digital signal is acquired (stimulable phosphor scanning), i.e., the subject matter of claims 5/11. It contains no interval division and no interval-varying smoothing.
• Anticipation conclusion: Does not anticipate any claim of the '325 patent. Different invention (scan-direction determination / read-out time reduction), not smoothing.

4. US 4,720,786 — "Method of compensating for offset distortion in rows of electrophoretic patterns"

• Full citation: U.S. Patent No. 4,720,786, inventor Makoto Hara (the same inventor as the '325 patent), assignee Fuji Photo Film Co., Ltd.; JP priority 60(1985)-…; see the '325 specification, which lists this as co-pending work.
• Dates: JP priority 1985-04-19; granted 1988-01-19.
• Description (directly retrieved): Detects at least two bands in the lower part of each resolved row, numbers the bands, derives a correlation (regression line/curve) between band number and migration distance, and corrects offset distortion between rows using the intercept difference. Notably, its specification expressly observes the same physical fact underlying the '325 invention: "there is a difference in the band spaces between the upper part and the lower part of the resolved pattern," with bands dense in the upper region.
• § 102 status: § 102(e) (US filing ~1986).
• Potential § 102 relevance: Same field, same inventor, same observation about upper/lower band density. But its claims are directed to distortion correction, not interval-variable smoothing.
• Anticipation conclusion: Does not anticipate claims 1 or 7. It supplies motivation/context (the upper-region band crowding that the '325 smoothing addresses), making it strong § 103 art, but it does not disclose the claimed interval-divided variable-mask smoothing.

5. US 4,802,101 — "Signal processing method for determining base sequence of nucleic acid"

• Full citation: U.S. Patent No. 4,802,101, assignee Fuji Photo Film Co., Ltd.; JP priority.
• Dates: JP priority 1985-08-19; granted 1989-01-31.
• Description: A Fuji Photo Film signal-processing method for determining nucleic-acid base sequence from autoradiograph digital signals. Caveat: I could not retrieve the full specification of this reference within the session step limit, so this description is at lower confidence than the others. Based on its title, assignee, priority era, and the '325 specification's cross-references to the inventor's co-pending applications (corresponding to U.S. Pat. No. 4,720,786 and applications 06/849,187, 06/866,355, 07/030,062), it appears to concern base-sequence determination with distortion/noise correction.
• § 102 status: § 102(e) (US filing before applicant's invention).
• Potential § 102 relevance: Same technical field and same assignee; may disclose band-detection/sequencing and distortion-correction steps. No indication (and no basis in the retrieved material) that it discloses interval division with interval-varying moving-average mask size.
• Anticipation conclusion: Does not anticipate claims 1 or 7 on the record available; flagged as unverified. Should be re-checked against its full text before relying on the negative.

Synthesis: what the cited art does and does not show

None of the five cited references, taken alone, anticipates claim 1 or claim 7. The reason is structural, not merely evidentiary:

• Claim 1 requires smoothing each interval "using a moving average filter having a mask size that is varied for each said interval."
• Claim 7 requires smoothing each interval "according to spaces of bands in each said interval."

The cited references collectively disclose: (i) one-dimensional waveform/graph preparation (US 4,777,597; US 4,837,733); (ii) smoothing by convolution with a single filter function (US 4,837,733); (iii) thresholding/histogram sampling-point detection (US 4,777,597; US 4,837,733); (iv) stimulable-phosphor read-out and scan-direction determination (US 4,665,312); and (v) recognition that upper-region band spacing is denser than lower-region spacing, with per-row distortion correction (US 4,720,786). What is absent from every one of them is the combination of (a) dividing the one-dimensional waveform into at least two intervals and (b) deliberately varying the smoothing strength/mask size between those intervals. That interval-adaptive smoothing is the novelty the '325 patent claims.

Consequently, the correct § 102 characterization of these references is negative: they are § 103 (obviousness) art and background, not § 102 anticipation art. Even the individual feature match-ups (stimulable-phosphor read-out → claims 5/11; radiosensitive-material read-out → claims 6/12; G/A/T/C four-group scheme → claims 4/10) cannot yield anticipation, because claims 4–6 and 10–12 are dependent and incorporate all limitations of claims 1 and 7.

Closest reference: US 4,837,733 (smoothing by convolution) is the single most relevant citation, and US 4,720,786 supplies the band-spacing observation; a § 103 challenge would most plausibly be built on these two in combination — but that is an obviousness question, not the § 102 question asked.

Caveats

1. No anticipation found is a technical conclusion about the cited art, not a legal opinion. A formal § 102 analysis would also require the exact U.S. filing dates of each reference (for their § 102(e) dates) and the applicant's actual invention date.
2. US 4,802,101's description is unverified — its full text was not retrieved; the negative anticipation result for it should be reconfirmed.
3. US 4,777,597 and US 4,837,733 claim 1983 JP priority, but under In re Hilmer their § 102(e) dates are their U.S. filing dates, not 1983 — I used the U.S. chain dates (07/024,909 era for the former; 06/568,906, filed 1984-01-06, for the latter).
4. Other potentially relevant art not on the face: the '325 specification itself references co-pending applications 07/161,248, 06/917,606, 06/917,609, 07/423,686, 07/378,509 and U.S. Pat. No. 4,777,597, plus EP 0199327 / EP 0242629 and JP 61-233369-type family members (shown in the "Similar Documents" list). If the goal is a complete § 102/§ 103 picture, those should be pulled in addition to the five face citations.

Sources: Google Patents full text, https://patents.google.com/patent/US4972325/en ; US 4,837,733 (https://patents.google.com/patent/US4837733 and https://patents.justia.com/patent/4837733); US 4,777,597 (https://patents.google.com/patent/[US4777597A](/patent/US4777597A)/en); US 4,665,312 (https://patents.justia.com/patent/4665312); US 4,720,786 (https://patents.google.com/patent/US4720786 and https://patents.justia.com/patent/4720786).

I'll ground this analysis in the actual cited prior art before reasoning about § 103.

Obviousness Analysis — U.S. Patent No. 4,972,325 under 35 U.S.C. § 103 (pre‑AIA)

Analyst note on scope. This builds on the previously generated bibliographic, claim‑overview, and litigation sections rather than repeating them. It uses the five references listed in the "Citations (5)" / "Patent Citations" section of the patent page — US 4,777,597; US 4,835,733; US 4,665,312; US 4,720,786; US 4,802,101 — plus the Japanese/EP publications that are counterparts of those same Fuji Photo Film disclosures. Where I rely on a family member rather than the exact U.S. patent (because the family member's publication date is what makes it available as art), I say so.

I. Bottom line

1. There is a strong, articulable § 103 case against claim 1 and claim 7, and against all dependent claims, built on the Shiraishi/Fuji smoothing art (US 4,777,597; US 4,835,733; their EP/JP counterparts) in view of the band‑spacing‑gradient art (US 4,720,786 / EP0199327), and further in view of the band‑detection‑and‑sequencing art (US 4,802,101).
2. The case is materially weakened by two threshold problems that any challenger must solve first, and which the previously generated sections did not surface:
   • Timing/"by another" problems with US 4,720,786 and US 4,802,101 (both name the same sole inventor, Makoto Hara, as the '325 patent, and both published after the '325 priority date of 1986‑03‑29).
   • Common‑ownership (§ 103(c)) problems with every cited reference, since all five are Fuji Photo Film assignees just like the '325 patent.
3. Net assessment: more likely than not invalid over the Shiraishi art + the known band‑spacing gradient, provided the challenger can either (a) characterize US 4,720,786 / US 4,802,101 as § 102(a)/(b) art, or (b) succeed in using them as evidence of the level of ordinary skill rather than as § 103 references, or (c) rely on non‑patent literature for the filtering principles. Without those, the case is noticeably thinner but still arguable.

II. Governing law and the "prior art" that is actually available

Framework: pre‑AIA § 103(a) (the application was filed 1987‑03‑30 and the patent is being analyzed under the law in force at its filing). A reference must first qualify as prior art under § 102 before it can be used in a § 103 combination. Graham v. John Deere Co., 383 U.S. 1 (1966) supplies the four factual inquiries; KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) supplies the rationales (later, but now applied to all patents, including pre‑AIA ones).

Qualification table

Deliberately excluded (important): Google Patents and the search results surface US 4,980,827 / JPS6285862A / JP60226092A (filed 1985‑10‑11, published 1987‑04‑20), which contains the single most on‑point teaching in the whole field: that the comparison/search window "σ … may be defined as a function of the peak position y_b … and it can be so set that σ becomes smaller toward the electrophoresis‑starting position" (see the description and claims at https://patents.justia.com/patent/[4980827](/patent/4980827)). That document is not available against '325 — its U.S. application was filed 1986‑10‑10 (after the 1986‑03‑29 priority date) and its JP publication is 1987‑04‑20 (after the priority date). Do not cite it as § 103 art; at most it is evidence of what the field was doing contemporaneously. The same caution applies to EP0242629 / JP 61‑69074, which teaches a threshold "fluctuating from interval to interval" — it is expressly identified in the '325 specification itself as co‑pending work (see the '325 text: "our co‑pending Japanese Patent Application Nos. … 61(1986)‑69074"), so it is an admission of common development, not prior art.

§ 103(c) flag. All five cited references are assigned to Fuji Photo Film Co., Ltd., as is '325. Section 103(c) disqualifies subject matter that qualifies as prior art only under § 102(e), (f) or (g) where the subject matter and the claimed invention were commonly owned at the time of invention. If the version of § 103(c) in force for a 1987 filing governs (the 1984 enactment, covering only § 102(f)/(g)), common ownership would not disqualify the § 102(e) Shiraishi patents. If the post‑1999 version (which added § 102(e)) governs, it would. I am not fully certain which version a court would apply to this application, and I flag this rather than resolve it. Note that the EP/JP publications are § 102(b) art in their own right and are therefore never disqualified by § 103(c) — which is why I anchor the combinations on the published EP/JP counterparts wherever the motivation depends on the teaching.

III. What the '325 claims actually add over the art (delta analysis)

Stripping the preamble and the intended‑use language, claim 1 requires only:

Claim 7 is broader in step (3) — any "smoothing means … according to spaces of bands in each said interval" — and adds the band‑detection and base‑determination steps. The specification itself concedes that the smoothing implements "various known noise reduction filters … a moving average filter, a weighted mean filter and a filter having low‑pass characteristics." That is an express admission that the filter is old; the only question is whether changing the filter's strength along the migration axis was obvious.

The critical admission is in the '325 specification's own Summary: "There is generally a tendency that bands exist at a high density with narrow spaces therebetween in the upper part … and that bands exist sparsely with wide spaces therebetween toward the lower part … When the upper part … is signal‑processed by smoothing on the same condition as that of the lower part thereof, peaks of signal levels become indistinct." That is a statement of the known problem, and US 4,720,786 says the same thing in the same words‑worth of substance: "there is a difference in the band spaces between the upper part and the lower part of the resolved pattern … the band spaces are dense in the upper region, to cause an error in the determination of the base sequence."

IV. The specific combinations

Combination A — Claim 1

US 4,777,597 (or US 4,835,733) in view of US 4,720,786, optionally further in view of US 4,802,101.

• Shiraishi '597/'733 supplies: autoradiograph → digital signal; one‑dimensional waveform (position vs. level) per row; smoothing by convolution with a filter function; threshold processing; detection of sampling points; and the base‑sequence end use. See claims and description at https://patentimages.storage.googleapis.com/79/dd/c5/65fbc543f5bbab/US4837733.pdf and EP0113677A3.
• Hara '786 supplies the reason to treat the axis non‑uniformly: it identifies the dense‑top / sparse‑bottom band‑spacing gradient and explicitly warns that uniform treatment in the dense upper region "cause[s] an error in the determination of the base sequence," and it advocates handling the upper and lower regions differently (regression of band number vs. migration distance is "nearly linear" in the lower part).
• Result: every element of claim 1 is taught or suggested. The only element not literally in either reference — "mask size varied for each interval" — is the predictable result of applying a known filter (convolution with a filter function) with a window that a PHOSITA would size to the local feature spacing.

Combination B — Claim 7

US 4,802,101 in view of US 4,777,597 (or US 4,835,733) and US 4,720,786.

• '101 gives the waveform‑per‑lane step and, critically, the closing step — comparing band positions across rows and "utilizing said position of said band to determine said base sequence of said nucleic acids."
• '597/'733 gives the smoothing step.
• '786 gives the "according to spaces of bands in each said interval" functional limitation, because it is the reference that teaches that the band spacing differs by region and that processing must account for it.
• Claim 7 is broader than claim 1 (no moving‑average limitation), so it is easier to invalidate: any smoothing whose strength tracks band spacing reads on it.

Combination C — Dependent claims

V. Why a PHOSITA would have been motivated to combine

The motivation is unusually clean for a § 103 case, and it does not depend on hindsight reconstruction of the applicant's own figure:

1. Same field, same problem, same pipeline. All five cited references concern automated extraction of DNA/RNA base sequence from an autoradiograph of resolved rows, and all address the same documented failure mode — the resolved pattern "is liable to cause various distortion and noise" that "causes an error in the determination of band positions (peak positions of signal level)."
2. Common assignee and a shared development program. All five are Fuji Photo Film references; the '325 specification cross‑references a whole series of co‑pending Fuji applications (JP 60‑74899, 60‑74900, 60‑85275, 60‑85276, 60‑111185, 60‑111186, 61‑69074) as implementing the same overall method. Combining related components of a single, coordinated signal‑processing pipeline is the paradigm of an obvious combination. KSR, 550 U.S. at 417 ("a patent composed of several elements is not proved obvious merely by demonstrating that each of its elements was, independently, known in the prior art" — but conversely, "if a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way, using the technique is obvious").
3. The problem the invention solves is stated in the prior art, not discovered by the applicant. US 4,720,786: the band spaces differ between the upper and lower parts of the pattern, the upper region is dense, and the band order there is hard to determine. A PHOSITA told "the feature spacing varies by roughly an order of magnitude along this axis, and a fixed window destroys peaks in the dense region" has a direct, specific reason to make the window position‑dependent.
4. The result is the predictable consequence of a well‑understood trade‑off. In digital filtering (well established by 1986), widening a boxcar kernel monotonically improves noise attenuation and monotonically degrades the ability to resolve closely spaced peaks. The spacing of bands is known (from the same art) to be dense at the electrophoresis‑starting end and sparse at the far end; therefore, matching window width to local feature spacing is the only direction the trade‑off points. This is a "predictable variation[] of prior art elements according to known methods" with "a reasonable expectation of success." KSR, 550 U.S. at 417, 421.
5. Finite, small, and testable set of solutions — "obvious to try." The specification's own embodiment is a five‑member ladder of point averages (5‑, 4‑, 3‑, 2‑ and 1‑point), i.e., a choice among four or five integers, evaluated by whether the peak survives. Where "there are a finite number of identified, predictable solutions, a person of ordinary skill has good reason to pursue the known options." KSR, 550 U.S. at 421; In re Aller (varying a parameter of a known process to obtain an optimum is routine optimization, not invention when the result is predictable).
6. Trivial implementation cost / no new hardware. The '325 circuit is a counter ("factor control 13") that flips a coefficient set every K = N/M samples and a look‑up table (FIG. 4). Nothing about the change required new instrumentation, new chemistry, or an unexpected insight — it is a software parameter change in an existing pipeline.
7. The prior art was not working in a different direction. Shiraishi's smoothing disclosure uses a fixed convolution kernel and says nothing steering away from a variable one. Silence is not a teaching away. In re Kahn, 441 F.3d 977, 990 (Fed. Cir. 2006) (the question is whether the art discourages the combination, not whether it mentions it).

VI. Rebuttal / objective‑indicia analysis

For the patentee, the plausible arguments are:

• No reference expressly discloses the varying mask size. True, and worth saying: none of the five cited references (as distinct from the unavailable US 4,980,827 line) says "use a smaller mask where bands are dense." The challenger must supply that step by reasoning about filter window vs. feature spacing. Pre‑KSR panels applying a strict "teaching, suggestion, motivation" test would have found this dispositive; post‑KSR, foreseeable parameter optimization supplies the missing link. This is the patentee's best argument, and it is a real one.
• Teaching away? I found no teaching away. To the contrary, the field's own later‑filed document (US 4,980,827) sets the comparison window σ as a decreasing function of the distance from the start — i.e., the field moved in precisely the claimed direction. That document is not art against '325, but it is powerful evidence that the claimed direction was the field's natural trajectory rather than an unexpected discovery.
• Objective indicia: I found no evidence in the materials available to me of unexpected results, commercial success attributable to the claimed feature, copying, licensing, or skepticism. The specification asserts no data showing that the ladder outperforms any alternative in an unexpected way; the described benefit (noise removed without flattening peaks) is exactly the expected effect of matching bandwidth to feature spacing. Note that Fuji's BAS/FLA‑series readers did achieve commercial significance in this era, but nothing in the record ties that success to the interval‑varying mask as opposed to, e.g., the stimulable phosphor sheet itself (which is separately claimed and separately old).
• Written description/claim‑construction defenses (e.g., whether "mask size" must be an integer pixel count, whether "interval" must be contiguous and cover the entire waveform) are available but narrow; claim 1's language is broad and functional‑adjacent, making a "gist" reading difficult to avoid.

VII. The two structural weaknesses in the invalidity case (flagged, not resolved)

1. Same‑inventor / timing problem with US 4,720,786 and US 4,802,101. Both name Makoto Hara as sole inventor — the same inventive entity as '325 — and neither published before 1986‑03‑29. If they are not "by another" for § 102(e) purposes, and their U.S. filings postdate the '325 invention date, they may not be § 103 references at all. If so, the motivation must come from the Shiraishi art plus general knowledge, or from the '325 specification's own background admission of the band‑density gradient — which the patentee has handed to the challenger in its own Summary. (The specification's admission is usable against the patentee as a concession about the state of the art.)
2. Common‑ownership / § 103(c). See § II above; the outcome turns on which version of § 103(c) governs, and I flag that I am not certain.

What would change the conclusion: (a) proof that US 4,720,786's U.S. filing date precedes 1986‑03‑29 or that it is otherwise available as § 102 art — I could not confirm its exact U.S. filing date from free sources; (b) location of a 1984–1985 journal or textbook teaching variable‑width smoothing of electropherograms (non‑patent literature would be § 102(b) art and immune to both problems above); or (c) the prosecution history, which I could not retrieve, showing the examiner's actual reason for allowance.

VIII. Confidence summary

IX. Cross‑reference and contradiction flags

• The previously generated "Patent summary" and "Litigation summary" sections are consistent with the claim text and dates I relied on here; I found no contradiction with them on the substance of § 103. Their observation that the patent is "Expired – Lifetime" with anticipated expiration 2007‑11‑20 is consistent with the Google Patents record and with the fee payments at years 4, 8 and 12.
• The date discrepancy those sections flagged (prompt date vs. a retrieved 2026‑09‑27 date) persists; the header on this task states April 26, 2026 while the fetched page reports a 2026‑09‑26T08:18Z retrieval. It does not affect any § 103 conclusion.
• The previously generated section rendered the assignee as the literal "FUJI PHTO FILM CO., LTD." I have not corrected that spelling and have used it only when quoting the assignment record; elsewhere I use "Fuji Photo Film Co., Ltd." as the party named in the references themselves. Claim 12's "radiosenstive" spelling, also noted earlier, is likewise left uncorrected.
• One thing I positively assert and the earlier sections did not cover: US 4,980,827 / JPS6285862A is not available prior art against '325 despite being the most on‑point document in the field. A challenger who cites it invites a fatal date objection. The same is true of EP0242629 / JP 61‑69074, which the '325 specification itself identifies as co‑pending work.

Primary sources cited: https://patents.google.com/patent/[US4972325](/patent/US4972325)/en (full text, claims 1–12); https://patents.google.com/patent/US4802101 ; https://patents.google.com/patent/US4720786 ; http://patentimages.storage.googleapis.com/7e/43/2a/ec4d4ffe35ea6b/EP0199327A3.pdf ; http://patentimages.storage.googleapis.com/7b/99/a1/b1114fbec9a7fa/EP0113677A3.pdf ; https://patentimages.storage.googleapis.com/79/dd/c5/65fbc543f5bbab/US4837733.pdf ; https://patents.google.com/patent/JPS59126280A/en ; https://patents.google.com/patent/JPS59126530A/en ; https://patents.google.com/patent/US4980827 (flagged as not prior art) ; https://patentimages.storage.googleapis.com/7b/52/ea/09c29bde8b9775/EP0242629A3.pdf (not prior art).