US 8556983

I'll search for authoritative information on US 8,556,983 and related litigation/docket records.

Let me verify the granted claim set and check for any litigation or Federal Circuit activity.

US 8,556,983 B2 — Summary

Sourcing note: I do not have direct terminal access to USPTO PatentCenter/PAIR, PACER, or the Federal Circuit docket system in this session, so the below is drawn from the patent's public full-text record and open web sources (Google Patents, FreePatentsOnline, uspto.report, Darts-IP) rather than a live docket query. Where I could not verify something, I flag it.

Bibliographic data

Caveat on dates: Google Patents lists the priority/prior-art date as 2001-05-25, inherited through the long continuation chain (10/160,667 → 10/305,652 → 10/724,010 → 11/002,573 → 11/671,745 → 12/660,529 → 13/044,413). The granted claims, however, are directed to minimum/maximum implant-thickness methods that appear tied to the 2009–2010 disclosures (e.g., 61/155,359, 12/660,529, 12/712,072). The effective priority date of the granted claims is a legal determination I cannot make here.

Abstract (as granted)

"Methods and devices are disclosed relating improved articular models, implant components, and related guide tools and procedures. In addition, methods and devices are disclosed relating articular models, implant components, and/or related guide tools and procedures that include one or more features derived from patient-data, for example, images of the patient's joint. The data can be used to create a model for analyzing a patient's joint and to devise and evaluate a course of corrective action. The data also can be used to create patient-adapted implant components and related tools and procedures."

Independent claims (plain language)

The granted patent has 18 claims, of which claims 1, 9, and 16 are independent. All three are method claims — notably, the granted claims are far narrower than the published application (US 2011/0295378 A1, which carried apparatus claims such as claim 52 to a femoral implant with patient-specific condylar curvatures). The applicant appears to have replaced the apparatus claims with method claims during prosecution.

Claim 1 — Method of selecting/designing a patient-matched articular implant

1. Preoperatively identify, from patient-specific data of the patient's joint, a curvature on an articular surface of the joint.
2. Select an outer (joint-facing) articular-surface curvature for the implant that substantially matches that identified curvature (or a predetermined percentage of it).
3. Select a desired minimum implant thickness.
4. Select and/or design one or more surface facets on the inner (bone-facing) surface that provide at least that minimum thickness between the facet(s) and the selected articular-surface curvature.

In short: measure the patient's joint curvature, match the implant's bearing surface to it, pick a minimum thickness, then derive bone-facing cut facets that preserve that thickness.

Claim 9 — Method of making an implant meeting a minimum thickness threshold

1. Identify a desired minimal thickness threshold.
2. Derive, from patient-specific joint data, a substantially coronal or substantially sagittal curvature for the implant's articular surface.
3. Modify that curvature so the implant falls within the minimum thickness threshold.

Claim 16 — Method of making an implant meeting a maximum thickness threshold

1. Identify a desired maximum thickness threshold.
2. Derive, from patient-specific joint data, a coronal or sagittal curvature for a condylar articular surface of the implant.
3. Modify that curvature so the implant falls within the maximum thickness threshold.

The key distinction between claims 1, 9, and 16 is the constraint being enforced: claim 1 = thickness floor while preserving a patient-matched curvature; claim 9 = minimum-thickness compliance via curvature modification; claim 16 = maximum-thickness (bone-preservation/"pre-primary" sizing) compliance via curvature modification.

Dependent claim map

• Implant type: 2 (femoral, tibial, patellar, hip, shoulder, spinal), 3 (femoral)
• Design/manufacture technique: 4 (CAD/CAM), 5 (altering a blank), 6 (smoothing a line/curve derived from the patient's surface)
• Thickness determination/analysis: 7 and 12 (FEA), 8 and 13 (load-bearing analysis), 14 (FEA for max thickness), 15 (load-bearing analysis for max thickness)
• Curvature modification: 10 and 17 (increasing the radius of curvature)
• Dual threshold: 11 (adding a maximum threshold to claim 9), 18 (adding a minimum threshold to claim 16)

Litigation / CAFC 2026 docket check

I found no evidence of any 2026 Federal Circuit docket, appeal, or decision specifically naming US 8,556,983. Please treat that as an absence of evidence, not proof of absence — I could not query PACER or the CAFC docket directly.

What I could confirm:

• Google Patents flags this patent as part of a "family has litigation" grouping (Darts-IP family 45022734, https://patents.darts-ip.com/?family=45022734).
• The known ConforMIS enforcement campaign against Smith & Nephew generated inter partes reviews on sibling patents in this family — not on the '983 patent itself. Examples found in the record: IPR2017-00373 (U.S. 8,551,169), IPR2017-00510 (U.S. 7,981,158), IPR2017-00778 (U.S. 8,062,302).
• I did not find any IPR petition, district court filing, or appeal in which US 8,556,983 is the asserted patent.

Practical significance: because the patent shows as Expired – Lifetime with anticipated expiration 2022-05-28, it can no longer be asserted for new infringement. Any 2026 Federal Circuit activity would therefore most plausibly be appellate review of an earlier judgment, an IPR appeal, or a case in which the '983 patent is cited as prior art or in a family/patent-marking context — not fresh infringement litigation on this patent.

Uncertainties I want to flag explicitly

1. Claim count. I have the claim set as claims 1–18 with independents 1, 9, 16 from FreePatentsOnline's rendering of the '983 grant. I was not able to cross-verify the exact granted claim count against a second independent source (e.g., PTAB's or the official printed patent's "Claims" section) in this session. If precision matters legally, verify against the official USPTO record.
2. Do not confuse the grant with the publication. US 2011/0295378 A1 contains apparatus claims (e.g., a femoral implant with a patient-specific condylar curvature in a first plane and an engineered/constant curvature in a second plane). Those are not the granted claims of the '983 patent.
3. Priority chain. The 2001-05-25 date is the earliest claimed benefit, but the granted method claims likely depend on later-filed subject matter.
4. No CAFC 2026 result is a negative finding only, and I did not search litigation databases (Darts-IP, Lex Machina, Docket Alarm) directly.

I'll search for litigation involving US patent 8556983 across patent litigation sources.

Let me search more specifically for known Conformis litigation and PTAB proceedings involving this patent.

Let me search for cases specifically asserting the '983 patent and check litigation portals.

Bottom line

Based on the litigation sources I was able to search, I could not identify any case in which U.S. Patent No. 8,556,983 ("the '983 patent") is specifically named as an asserted patent. I also found no IPR/PGR specifically challenging the '983 patent.

What I did find is that the '983 patent belongs to a Conformis patent family that has been extensively litigated, and that Conformis (the assignee) ran a long series of patent campaigns against orthopedic competitors. But in every case record I retrieved, the patents asserted were other members of the family — not the '983 patent. I want to be explicit about that distinction rather than imply the '983 patent was asserted.

I was unable to directly open the Unified Patents litigation case list, PACER, or Docket Alarm/RPX (my search tool returned a step-limit message), so I cannot rule out that the '983 patent appears in a docket I could not reach. Treat the "no litigation found" conclusion as limited to the sources actually searched.

What the sources show about the '983 patent's family

• Google Patents' page for US 8,556,983 B2 carries a "Family has litigation / First worldwide family litigation filed" badge linking to a Darts-IP family record (family=45022734). (https://patents.google.com/patent/US8556983/en). This tells us the family has been litigated; it does not establish that the '983 patent itself was asserted.
• Assignee: Conformis, Inc. Inventors: Bojarski, Chao, Slamin, Lang, Fitz, Steines. Priority date 2001-05-25; filed 2011-03-09; granted 2013-10-15; status listed as "Expired – Lifetime."

Conformis patent litigation I found (none of which lists the '983 patent)

Patents actually reported as asserted in these cases (as reflected in the complaints/summaries I retrieved) include, among others: U.S. 8,377,129; 8,568,480-type/'304; '482; '161; 7,799,077; 8,077,950; 8,638,998; 9,180,015; and various 10/xxxxxxx patents in the DePuy Synthes action. The Conformis complaint excerpts I retrieved recite patents such as the '129, '304, '161, '482, '539, '079, '077, '950, '998, and '015 patents — not the '983 patent.

Also note a separate case where Conformis was the defendant: Bonutti Skeletal Innovations LLC v. Conformis (D. Del., e.g., 12-cv-1107/1109/1110/1111), involving Bonutti's knee-surgery patents — unrelated to the '983 patent.

Important caveats

1. The Knobbe Martens litigation slide deck (2021) summarizing the history of Conformis's campaigns refers to a "'953 patent" in the Smith & Nephew matter (IPR filed Mar 2017; Conformis voluntarily dismissed the '953 patent Feb 2017). I am not treating that as the '983 patent. "953" ≠ "983," and I have no basis to assume a typo. I flag it only so you are aware a similar-looking number appears in that record.
2. I could not confirm whether the '983 patent was among the nine patents asserted against Smith & Nephew, or among its 16 IPRs, because the source I retrieved lists only partial patent identifiers. This is the most plausible place the '983 patent could have surfaced, so it is worth pulling the Smith & Nephew complaints and IPR petitions (D. Mass. 1:16-cv-10420; PTAB IPR2016–2017 Conformis filings) directly to verify.
3. No IPR, PGR, or CBM specifically naming US 8,556,983 appeared in my searches.

Recommended verification steps

• PTAB (USPTO Patent Trial and Appeal Board) — search "Conformis" as patent owner and cross-check the '983 patent number directly at the PTAB End-to-End / IPR databases.
• Unified Patents litigation portal (https://portal.unifiedpatents.com/litigation) and RPX Insight — look up the patent by number under "asserted patents."
• PACER / Docket Alarm / CourtListener — pull the full complaints in the cases above (especially the Smith & Nephew and Wright/Tornier matters) and confirm the asserted-patent list.

If you'd like, I can drill into the Smith & Nephew and Wright complaints and the corresponding PTAB petitions to determine definitively whether the '983 patent was ever asserted — that is the concrete next step given the ambiguity in the records I retrieved.

I'll start by checking the structured ODP block against public sources, then search for any PTAB activity, related litigation, and Federal Circuit appeals.

Let me run targeted searches on Conformis PTAB activity and the known litigation campaigns that might have generated IPRs on this family.

Let me search for a specific Smith & Nephew IPR against the '983 patent and check the Conformis IPR campaign details.

Proceedings overview

Zero AIA trial proceedings on file for US 8,556,983. The structured "PTAB proceedings on file" block (USPTO Open Data Portal) reports no AIA trials — no IPR, PGR, or CBM — and independent public searching corroborates that: I found no petition, institution decision, FWD, or appeal naming U.S. Patent No. 8,556,983. The breakdown is therefore 0 active / 0 claims invalidated / 0 sustained / 0 settled / 0 institution-denied — the patent itself has never been before the Board.

Bottom-line defensive posture: this is not a "hardened-by-surviving-IPR" story and not a "claims already canceled" story. It is a "never-tested patent" story. No petitioner has ever put 8,556,983 at risk, so every claim (1 through the last dependent claim) is UNTESTED and stands as issued — but there is also no § 315(e)(2) estoppel protecting a defendant, meaning the prior-art field is wide open. Two facts dominate the practical calculus:

1. The patent is expired. Google Patents records the anticipated expiration as 2022-05-28 and the legal status as "Expired – Lifetime." Assertion today is a damages-only proposition covering the pre-expiration window (subject to § 286's six-year lookback), not an injunction threat.
2. It was never among the Conformis patents asserted in the campaign that did generate IPRs. The Smith & Nephew IPR wave (below) hit sibling Conformis patents, not this one. That is the single best explanation for the empty docket.

Proceedings

No IPR, PGR, or CBM proceeding exists against U.S. Patent No. 8,556,983. Because the task calls for a per-proceeding walkthrough, below is (A) the negative confirmation, and (B) the adjacent proceeding cluster — the Conformis-family IPRs that a defendant will inevitably be told about — flagged clearly as not on this patent.

No proceeding — US 8,556,983 has never been challenged at the PTAB

• Type: n/a
• Filed: n/a
• Status: No AIA trial proceedings on file (ODP). Corroborated by public search.
• Judge panel: n/a
• Petition grounds: n/a
• Institution decision: n/a
• Final Written Decision: n/a
• Settlement / termination: n/a
• Appeal: n/a
• Defensive value: Nothing has been canceled and nothing has been sustained by the Board. A defendant cannot point to an FWD binding the patent owner, and equally cannot claim any claim is dead. The entire claim set is live-but-expired and never adjudicated.

Context — Smith & Nephew v. ConforMIS, the family IPR campaign (NOT on 8,556,983)

These are the proceedings a defendant will find when researching "Conformis PTAB" and mistaking them for this patent. Per Knobbe Martens' own litigation history (https://www.knobbe.com/wp-content/uploads/2021/06/Developing-Effective-Patent-Strategies-for-Personalized-Medical-Devices-in-Europe-and-the-US.pdf), following ConforMIS, Inc. v. Smith & Nephew, Inc., No. 1:16-cv-10420-IT (D. Mass., filed 2016-02-29), Smith & Nephew filed 16 IPRs against 9 Conformis patents; the litigation was stayed 2017-04; FWDs held claims unpatentable in four of those patents; and the parties settled in September 2018.

Representative proceedings I could verify by name (each on a different patent):

• IPR2017-00373 — Smith & Nephew, Inc. v. ConforMIS, Inc., U.S. Patent No. 8,551,169. FWD 2018-06-12 holding claims 29 and 30 unpatentable under § 103 (Ground 1: CAOS + Radermacher + Woolson; Ground 2: Swaelens + Woolson). ConforMIS filed a Notice of Appeal on 2018-08-13 (35 U.S.C. §§ 142/319), raising Graham/motivation-to-combine, reply-evidence, APA, and substantial-evidence issues. Panel: Scanlon, Bunting, Worth, Wieker.
• IPR2017-00511 — Smith & Nephew v. ConforMIS, U.S. Patent No. 7,981,158. FWD 2018-06-11 holding claims 66–72 and 81 unpatentable under § 103 (CAOS + Woolson + Alexander). Note the SAS order of 2018-04-27 (§ 42.5) modifying institution to all challenged claims, followed by a joint motion to limit back to 66–72 and 81.
• IPR2017-00510 and IPR2017-00544 (U.S. Pat. Nos. 7,981,158 and 7,534,263, respectively) — consolidated for a 2018-03-13 oral hearing.
• IPR2017-00778 / -00779 / -00780 — three petitions against U.S. Patent No. 8,062,302.
• IPR2017-00487 / -00488 (U.S. Pat. No. 9,295,482) and IPR2017-00984 (a Smith & Nephew PLC petition; institution denied).

Defensive value of this cluster for your case: it confirms (a) the Conformis portfolio was aggressively IPR'd by a well-funded, sophisticated petitioner using a consistent Radermacher/Alexander/Woolson/CAOS art stack, (b) Smith & Nephew won FWDs on several patents, and (c) 8,556,983 was not among them. A defendant cannot borrow those judgments — PTAB estoppel and cancellation run patent-by-patent.

⚠️ Unverified lead — do not rely on it. One Knobbe slide deck references a "953 patent" that ConforMIS voluntarily dismissed in February 2017, with a "953 IPR" filed March 2017 and instituted. That shorthand is not "983," and I could not confirm which patent it denotes. Treat any claim that "IPR2017-xxxxx invalidated 8,556,983" as false unless someone produces the proceeding number from PTAB E2E. I found no such proceeding.

Strategic summary

Claim status. All claims of 8,556,983 are UNTESTED. None are canceled; none have been confirmed by the Board. The patent issued 2013-10-15 from Application No. 13/044,413 (filed 2011-03-09, priority 2001-05-25) and its legal status is Expired – Lifetime with anticipated expiration 2022-05-28. Because a plaintiff asserting an expired patent cannot recover post-expiration damages and faces § 286's six-year bar on the front end, the realistic recovery window is narrow. That, not any PTAB outcome, is the primary defensive asset here.

Estoppel landscape. There is no estoppel. No petitioner has ever been a party to an IPR/PGR on this patent, so § 315(e)(2) does not bar anyone. A defendant today may raise in district court any § 102/§ 103/§ 112 defense, and may file its own IPR. Two timing cautions: (i) § 315(b) gives a petitioner one year from service of an infringement complaint to file an IPR — and that window is jurisdictional-adjacent and unforgiving; (ii) the expired status does not itself bar IPR, but the Board increasingly weighs settled expectations and workload in discretionary-denial practice, so expiring patents are weaker IPR candidates. Critically, the prior-art stack that defeated the sibling patents — Radermacher (WO 93/25157), Alexander (WO 00/35346), Woolson (U.S. 4,841,975), the Radermacher CAOS article, and Swaelens — has never been applied to 8,556,983. That is the most obvious unexplored ground, and it cost Smith & Nephew nothing to assemble.

Pattern signals. The petitioner pattern is a single large defendant filing a portfolio-wide salvo (S&N, 16 IPRs / 9 patents, 2016–2017), not a serial filer targeting one patent and not a defensive aggregator — I found no evidence of Unified Patents, RPX, or any similar entity in the chain for this patent. The patent owner did appeal adverse FWDs in the campaign (e.g., the 2018-08-13 notice in IPR2017-00373), but the September 2018 global settlement likely mooted those appeals; I could not verify the Federal Circuit dispositions and will not guess at docket numbers. Beyond AIA trials, Conformis also litigated portfolios against Zimmer, MicroPort, Wright Medical/Tornier (D. Del. 1:20-cv-00562, filed 2020-04-24), and Exactech, and in 2012 filed an interference against an Imperial Innovations application — none of which is an AIA trial.

Recommended next steps

1. Say plainly, in writing, that there is no PTAB activity. If opposing counsel or a demand letter implies that 8,556,983 "survived IPRs" or was "upheld by the PTAB," that is unsupported — no such proceeding exists (verify live at PTAB E2E: https://ptacts.uspto.gov/ and the PTAB Decisions site: https://www.uspto.gov/patents/patent-trial-and-appeal-board/ptab-decisions).
2. Do not cite sibling FWDs as if they bind this patent. You may cite the reasoning as persuasive authority and for claim-construction insight, but cancellation and estoppel do not cross patents. The two FWDs worth reading are IPR2017-00778/-779/-780 (U.S. 8,062,302) and IPR2017-00511 (U.S. 7,981,158) — both built on CAOS + Radermacher + Alexander + Woolson. CourtListener and the docketalarm mirror at https://www.docketalarm.com/cases/PTAB/IPR2017-00544/ host them.
3. Check the real threat vector — expiration and litigation posture, not validity. Confirm the actual expiration date including any PTA from the face of the patent and the Patent Center record; Google Patents' 2022-05-28 is an assumption it labels as such. If the patent expired, scope any exposure analysis to pre-expiration sales/uses and the § 286 window.
4. If you are served and want to invalidate, mind the clock. If a complaint asserting 8,556,983 is served, the § 315(b) one-year clock starts on service. Because no estoppel exists, you retain the full Radermacher/Alexander/Woolson/CAOS/Swaelens stack, plus everything in the ~1,600-reference prosecution history — but weigh any IPR against the patent's expired status and the Board's discretionary-denial practice.
5. If you are the patent owner's adversary on the portfolio level, note that the settlement-created FWDs are the real precedent; and if you are considering asserting this patent, expect the Symmetry/Conformis-family validity fight the PTAB has never had.

Confidence statement: High confidence that no AIA trial proceeding exists on U.S. Patent No. 8,556,983 as of this analysis. Medium confidence in the enumeration of the adjacent S&N proceedings' patent numbers (assembled from public filings and secondary sources, not from the ODP). I did not invent any proceeding number, and I have flagged the one ambiguous "953" reference rather than resolving it.

I'll research the assignment record for US 8556983, starting with the records surfaced in the patent text and cross-checking against USPTO, SEC filings, and news coverage.

Let me dig into the specific assignment records, reel/frame numbers, and correspondent details.

US 8,556,983 — Ownership & Assignment Reconstruction

Important data limitation up front: I was able to retrieve the substance of this patent's recorded post-issuance assignments (conveyance type, assignor, assignee, and event dates) from the patent's own prosecution/legal-events record, but I could not retrieve the reel/frame numbers or the correspondent-of-record for the individual recordings from the sources available to me. The USPTO Assignment Center would not render its per-record detail into my searches. Accordingly, no reel/frame numbers or correspondent names appear below — I will not invent them. Where a field is unavailable I say so explicitly rather than filling it in. Verification link at the end.

Inventors

Basis: the 2011-08-17 recorded assignment ("ASSIGNMENT OF ASSIGNEE INTEREST") lists all six inventors as assignors and CONFORMIS, INC. as assignee. Conformis was founded in 2004 by Philipp Lang, M.D. (Orthopedics This Week, Sept. 2023), and Lang's termination of employment is documented indirectly — Conformis's SEC filings state that revenue-share obligations to Dr. Lang "survived the termination of Dr. Lang's employment with us." So at least one named inventor (Lang) had departed the original assignee by the time of the 2018–2019 filings.

Unusual-pattern check: The "all inventors gone within 12 months of filing" fire-sale precursor is not observed here in the classic form, but there are two structural oddities worth flagging:

1. Inexact provenance of the priority chain. This application (US 13/044,413, filed 2011-03-09) is a continuation claiming priority to 2001-05-25. Conformis was not incorporated until June 2004 — i.e., the earliest priority filings predate the original assignee. Conformis's own 10-K states that its intangibles were "licensed from ImaTx as part of the spin-out transaction in 2004." So the earliest priority applications almost certainly originated with Imatx, Inc., the predecessor/spin-out parent, not Conformis. I could not verify the original applicant of record on the 2001–2003 applications, so treat this as a lead, not a finding.
2. Imatx, Inc. reappears as an assignor on the 2019 and 2021 security-interest recordings alongside Conformis and Conformis Cares LLC — consistent with it being a wholly owned subsidiary in the Conformis consolidated group, not an unrelated entity.

Original assignee

Conformis, Inc. (Billerica, Massachusetts; NASDAQ: CFMS until Sept. 2023).

• Product embodying the claims: Yes, unambiguously. Conformis sold patient-specific/patient-adapted knee and hip implants — iUni, iDuo, iTotal CR, iTotal PS, Identity Imprint, the Cordera and Actera hip systems. Orthopedics This Week reports more than 149,000 individually sized/shaped knee implants sold (123,000+ total, 26,000+ partial) over 19 years. This is a genuine operating company with real product revenue ($99.9M in 2021, $62.0M in 2022).
• Primary line of business: Design/manufacture of patient-adapted ("image-to-implant") orthopedic implants and single-use, patient-specific instrumentation.
• Current status: Acquired — not dissolved, not in bankruptcy. On 2023-09-05, restor3d, Inc. (Durham, NC) completed the merger contemplated by the June 22, 2023 Agreement and Plan of Merger; Conformis survived as a wholly owned subsidiary of restor3d, and CFMS ceased trading on Nasdaq. Price was $2.27/share cash (~$17M enterprise purchase). The company was under "going concern" doubt and self-described as "fighting for its very survival," but it went through a merger/stock acquisition, not a Chapter 7 or Chapter 11 proceeding. Shares outstanding at the 2023-08-31 special meeting: 7,878,332.

Assignment timeline

The following are the recorded post-issuance events on this patent. Reel/frame and correspondent fields were not retrievable and are omitted rather than fabricated. Execution dates are generally not exposed for these records; the dates below are the recorded/effective dates.

• Recorded 2011-08-17 — Reel not retrieved

  • Conveyance: Assignment of assignee interest (inventor → company)
  • Assignor: FITZ, WOLFGANG; CHAO, NAM; LANG, PHILIPP; BOJARSKI, RAYMOND A.; SLAMIN, JOHN; STEINES, DANIEL
  • Assignee: CONFORMIS, INC.
  • Correspondent: not retrieved
  • Context: Ordinary founder/inventor-to-company assignment perfecting title to the original operating assignee.

• Recorded 2018-07-31 — Reel not retrieved

  • Conveyance: Security Agreement (security interest)
  • Assignor: CONFORMIS, INC.
  • Assignee: OXFORD FINANCE LLC, AS COLLATERAL AGENT
  • Correspondent: not retrieved
  • Context: Securitization — collateral grant under Conformis's senior secured lending facility (the Oxford 2017/2018 loan and security agreement). Collateral, not a sale.

• Recorded 2019-06-25 — Reel not retrieved

  • Conveyance: Security Agreement (security interest)
  • Assignor: CONFORMIS CARES LLC; CONFORMIS, INC.; IMATX, INC.
  • Assignee: INNOVATUS LIFE SCIENCES LENDING FUND I, LP, AS COLLATERAL AGENT
  • Correspondent: not retrieved
  • Context: Securitization — replacement/refinancing lender's blanket IP lien covering the Conformis consolidated group.

• Recorded 2019-06-26 — Reel not retrieved

  • Conveyance: Release by Secured Party
  • Assignor: OXFORD FINANCE LLC
  • Assignee: CONFORMIS, INC.
  • Correspondent: not retrieved
  • Context: Release of the Oxford lien on payoff/refinancing by Innovatus — one day after the Innovatus grant, i.e., a same-day financing swap, not a transfer of ownership.

• Recorded 2021-11-22 — Reel not retrieved

  • Conveyance: Security Agreement (security interest)
  • Assignor: CONFORMIS, INC.; IMATX, INC.
  • Assignee: MIDCAP FINANCIAL TRUST, AS AGENT
  • Correspondent: not retrieved
  • Context: Securitization — third successive secured lender taking a blanket IP lien.

• Recorded 2021-11-23 — Reel not retrieved

  • Conveyance: Release by Secured Party
  • Assignor: CONFORMIS CARES LLC; CONFORMIS, INC.; IMATX, INC.
  • Assignee: INNOVATUS LIFE SCIENCES LENDING FUND I, LP
  • Correspondent: not retrieved
  • Context: Release of the Innovatus lien following the MidCap refinancing recorded the day before — again a financing swap, one day apart.

• Recorded 2023-09-06 — Reel not retrieved

  • Conveyance: Release by Secured Party
  • Assignor: MIDCAP FINANCIAL TRUST, AS AGENT FOR LENDERS
  • Assignee: IMATX, INC.; CONFORMIS, INC.
  • Correspondent: not retrieved
  • Context: Release of the MidCap lien one day after the restor3d merger closed (2023-09-05) — debt extinguished in the merger, clean title restored to the Conformis subsidiary.

No assignment of the patent itself to restor3d, Inc. is recorded. Title sits with Conformis, Inc., now a wholly owned subsidiary of restor3d. The absence of a separate restor3d recording is expected in a reverse-triangular merger, where the target survives.

Timeline diagram

timeline
    title Ownership of US 8556983
    2001 : Earliest priority filing
    2004 : Conformis spun out of Imatx
    2011 : Application filed as continuation
         : Inventors assign to Conformis Inc
    2013 : Patent issued
    2018 : Security interest to Oxford Finance
    2019 : Innovatus security interest recorded
         : Oxford release recorded
    2021 : MidCap security interest recorded
         : Innovatus release recorded
    2022 : Patent term expires
    2023 : MidCap release recorded
         : Conformis acquired by restor3d

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. The only true assignment is inventor → Conformis, Inc. (2011-08-17). Every other recorded event is a security interest or a release of one (2018-07-31; 2019-06-25; 2019-06-26; 2021-11-22; 2021-11-23; 2023-09-06). None moves ownership to a licensing-only vehicle. Imatx, Inc. and Conformis Cares LLC appear only as assignors of collateral, which is inconsistent with their being shell assignees. No "IP/Holdings/Ventures" successor appears anywhere in the chain.

2. Known asserter in the chain — NOT PRESENT. No Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, or Spangenberg entity appears as a recorded assignee. The secured lenders (Oxford Finance, Innovatus Life Sciences Lending Fund I, MidCap Financial Trust) are commercial lenders, a category distinct from NPEs; each has a public, mainstream venture-lending business.

3. Repeat correspondent across the chain — UNVERIFIABLE / INSUFFICIENT DATA. This is the one signal I was specifically asked to capture and could not — no correspondent-of-record names were retrievable. I note only that the same lender-side counsel would be the natural repeat player across the 2018/2019/2021/2023 lien recordings, but without the recorded correspondent fields I cannot assert recurrence. Flag this for manual pull from Assignment Center.

4. Cascading transfers — NOT PRESENT (as an NPE signal). There are tightly spaced pairs — Innovatus grant 2019-06-25 / Oxford release 2019-06-26, and MidCap grant 2021-11-22 / Innovatus release 2021-11-23 — but these are one-day financing refinancings, not chains of LLC-to-LLC conveyances. No shared anonymous address, no common-principal pattern, no succession of single-purpose assignees. Each pair nets to the same operating owner with a new lienholder.

5. Pre-litigation transfer — NOT PRESENT. Conformis's own enforcement suits — Wright Medical / Tornier Blueprint PSI shoulder (D. Del.), settled with Stryker/Wright Medical for a one-time $15M payment received 2021-10-15; Medacta (settled 2022-11-08, payment received 2022-12-12); a Paragon 28 non-exclusive license (2021-04-08); and an Aetna action — were all brought by the operating owner itself, not by an assignee that acquired the patent shortly before filing. No assignment precedes any of these suits.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7 or Chapter 11 filing by Conformis is evidenced. The 2023 exit was a merger under an Agreement and Plan of Merger with a 96% premium to market. It was a distressed, deeply discounted sale ($2.27/share, ~$17M) — arguably in the spirit of a fire-sale and worth noting — but it was not a bankruptcy proceeding, and the patents were not sold at auction out of an estate.

7. Privateering — NOT PRESENT. No operating company transferred the patent to a third-party NPE to assert on its behalf. Assertions were made in Conformis's own name against actual competitors (Wright Medical, Medacta) in the same market.

8. Defensive aggregator — NOT PRESENT. The chain terminates at restor3d, Inc., an operating orthopedic implant manufacturer, not at RPX, AST, LOT, Unified Patents, or OIN.

Additional fact bearing on any current assertion analysis: the patent's anticipated expiration is 2022-05-28 and its Google Patents legal status is "Expired – Lifetime." Whatever assertion value it had, it was exhausted roughly 14 months before the restor3d acquisition. This is material context for anyone assessing present-day assertion risk.

Verdict

Operating-company assertion.

Justification: every recorded conveyance other than the 2011-08-17 founder assignment is a security interest or its release (Oxford 2018-07-31; Innovatus 2019-06-25 with Oxford release 2019-06-26; MidCap 2021-11-22 with Innovatus release 2021-11-23; MidCap release 2023-09-06), and there is no assignment to any NPE or shell entity at any point. The original assignee, Conformis, Inc., was a real manufacturer that sold 149,000+ implants embodying the claims and sued genuine competitors in its own name (Wright Medical/Tornier Blueprint; Medacta; Aetna), recovering $15M from the Stryker/Wright settlement. Ownership now rests with Conformis, Inc. as a wholly owned subsidiary of restor3d, Inc. following the 2023-09-05 merger — an operating acquirer, not an asserter.

Caveats to this verdict: (a) the correspondent-of-record fields and reel/frame numbers were not obtainable from my accessible sources, so signal #3 is genuinely unresolved and should be pulled manually before any final opinion; and (b) the patent expired 2022-05-28, so it is unassertable today regardless of chain of title.

Verify at: USPTO Assignment Center, https://assignmentcenter.uspto.gov/ (search patent 8556983); mirror index at https://assignment.uspto.gov/patent/index.html. Cross-check the Google Patents legal-events tab at https://patents.google.com/patent/US8556983/en, which is the source of the dates recorded above (Google Patents does not expose reel/frame, which is why those remain open).

I'll research the cited prior-art references for US 8,556,983. Let me run several targeted searches.

Let me pull the actual "References Cited" list for this specific patent.

Prior-art analysis for US 8,556,983 B2

Bottom line up front: I could not retrieve a complete, verified front-page "(56) References Cited" list for this patent in this session, and I did not have PAIR/PatentCenter or the Image File Wrapper (the PTO‑892 forms that show which references the examiner actually applied). What follows is therefore (a) the subset of cited references I could verify against a rendering of the 8,556,983 grant, (b) a §102 screening of those references against the granted claim set as characterised in the earlier section of this analysis, and (c) explicit flagging of the many places where I am inferring rather than verifying. I have not invented any reference, number, or date below.

1. Why the "each citation → §102" exercise is harder — and narrower — than it looks

Three threshold points control the answer, and they all cut against finding anticipation:

(a) The granted claims are method claims about thickness constraints, not implant structures.
Per the previously generated section, the grant is claims 1–18 with independents 1, 9, and 16 — all methods that (i) derive a curvature from patient-specific data, (ii) match or modify the implant's articular-surface curvature, and (iii) enforce a minimum (claims 1, 9) or maximum (claim 16) implant thickness threshold, with dependents reciting CAD/CAM (4), blank alteration (5), curve smoothing (6), FEA (7, 12, 14), load-bearing analysis (8, 13, 15), and increasing radius of curvature (10, 17). This is the claim set I screened against.

(b) Anticipation requires every element in a single reference, arranged as claimed. Most of the cited art here is mid-20th-century mechanical prosthesis art (hip stems, meniscal bearings, bone cements, porous coatings). Structural implant references cannot anticipate a method reciting preoperative derivation of a patient-specific curvature plus a thickness-threshold-enforcing bone-facing facet design. At most they are §103 fodder.

(c) The effective priority date determines whether a reference is even §102 art. The application (13/044,413) was filed 9 March 2011 (pre-AIA §102 applies). Its earliest claimed benefit runs to 2001-05-25 through the continuation chain, but — as flagged in the earlier section — the thickness-threshold subject matter appears tied to the 2009–2010 disclosures (12/660,529; 12/712,072; 61/208,440/444; 61/339,766). If the granted claims get only a 2010 effective date, every cited reference published before ~March 2009 is pre-AIA §102(b) art and references published 2009–2010 are §102(a)/(e) art. If the claims somehow ride the 2001 date, then all references published on/after 2001-05-25 drop out of §102(b) and the 2001–2010 references only matter as §102(a)/(e) art. I cannot make that priority determination here, and it materially changes the answer.

(d) The patent is expired (anticipated expiration 2022-05-28). A §102 analysis is now relevant only to past damages/validity in a case covering pre-expiration conduct, or as context for the parallel family IPRs.

2. What I actually verified — the cited U.S. patent documents

The following U.S. patent documents appear in the "(56) References Cited" rendering of the 8,556,983 grant at uspto.report (https://uspto.report/patent/grant/8556983). Important caveat: the retrieved excerpt begins mid-list at 5,871,542; the list almost certainly extends both earlier (the family's citation lists typically open in the 3,314,420 / 3,605,123 range) and later (the excerpt cut off at 6,235,060). I am reporting the window I could verify and not backfilling from sibling patents, because the family lists are similar but not identical.

2A. CAD/CAM, imaging, and patient-specific design references — the only category with real §102 traction

Note on the identifier: the verified list includes 6,013,103 | Jan. 11, 2000 | Kaufman et al. (a prosthetic device reference also cited inside the '983 specification). I have kept it literally as retrieved; do not silently rewrite it.

2B. Instrumentation / resection-guide references — moderate relevance to dependents, none to the independents

2C. Implant structure references — §102 relevance essentially nil for the granted method claims

§102 conclusion for 2C: none of these discloses the combination of (i) preoperative derivation of an articular-surface curvature from patient-specific data, (ii) selection of a joint-facing curvature substantially matching it, and (iii) selection of a minimum/maximum implant thickness enforced through bone-facing facet geometry. They cannot anticipate claims 1, 9, or 16, nor their dependents.

3. Claim-by-claim §102 screen (building on the claim map in the earlier section)

Net: on the verified cited art, I find no single-reference §102 anticipation of any granted claim. That is a preliminary screen, not a legal opinion.

4. The genuinely dangerous art is not in the (56) list — it's the family's own art and the IPR record

The earlier section correctly notes no IPR targets the '983 patent itself. But the art that actually took down siblings is instructive and, for §103 (not §102), is the art to worry about:

• Radermacher WO 93/25157 (published 23 Dec. 1993) — individual templates from 3‑D reconstructions; used in IPR2017‑00115 (US 9,216,025) and IPR2017‑00778/00779 (US 8,062,302).
• Alexander WO 00/35346 (published 22 Jun. 2001) and WO 02/22014 — cartilage imaging/measurement, also in the IPR exhibits.
• Woolson US 4,841,975 (filed 15 Apr. 1987; issued 27 Jun. 1989).
• Fell WO 00/59411 (published 12 Oct. 2000) and Biscup US 2004/0117015 A1 (filed 5 Oct. 2001; published 17 Jun. 2004).
• Kenna US 4,646,729; Dunn US 4,759,350; Hofmann (1991); Carignan US 6,712,856; Schuster US 6,510,334; Swaelens WO 95/28688; Hemmy US 5,098,383; Pinczewski WO 01/66021.

Source: Smith & Nephew v. ConforMIS, IPR2017‑00115, Final Written Decision, Paper 33 (Apr. 19, 2018) — https://patentimages.storage.googleapis.com/.../Final-Written-Decision-8.pdf; and IPR2017‑00779 petition (Exhibit list) — https://patentdocs.typepad.com/files/smithnephew_ipr2017-00779_petition.pdf.

Note the structural point: those were §103 grounds built on multi-reference combinations. IPR2017‑00115 found claims 1–20 of the '025 patent unpatentable precisely through Rademacher + Alexander + Woolson (+ Biscup), i.e., not single-reference §102.

Non-patent literature cited on the family (also appearing in the '983-family record) is where the "CAD/CAM from patient data" concepts live: Radermacher CAOS (Clin. Orthop. 354:28, 1998); Radermacher 1994 MRCAS; Schiffers (Der Orthopäde 29:636, Jul. 2000); Portheine (2000, 2004); Hafez (2004, 2006); Chelule (ISTA 2002); Thoma (Der Orthopäde 29:641, Jul. 2000); Birnbaum (Spine 26(4):365, Feb. 2001); Tsai (IEEE, May 2001); Cohen (MICCAI 1998). These are the references I would pull first for a §102/§103 mapping against claims 1, 9, 16 — but each still needs to be checked for a thickness-threshold teaching.

5. Explicit contradictions / corrections to flag

1. Do not conflate citation lists across family members. During this search I encountered the "(56)" lists of US 8,568,480 and US 9,681,956 (ConforMIS siblings). Those lists are similar but not identical to 8,556,983's. I used only the list rendering for 8,556,983 (uspto.report/patent/grant/8556983) in §2 above, and reported it as a partial window.
2. A garbled entry I saw in a different patent's list — "6,791,188 B1 | 12/2005 | Christensen" — belongs to US 9,681,956's list, not 8,556,983's, and I have excluded it. (It also appears out of numeric order in that rendering, which is a further reason to distrust raw scrapes.)
3. "Cited" ≠ "applied." Nothing in §2 is a representation of what the examiner rejected over. The actual grounds would be in the PTO‑892/Office Actions in the 13/044,413 file wrapper, which I could not open.
4. The earlier section's priority caveat is the single biggest driver here. If the granted claims do not get the 2001 date, the 2001–2010 cited art becomes fully available §102(b) art; if they do, most of it is not §102(b) art at all. I have flagged both branches rather than pick one.

6. How to close the gaps (concrete next steps)

1. USPTO PatentCenter / PAIR, Application 13/044,413 → download the PTO‑892 and every Office Action. That gives the examiner-applied subset, which is the only subset with real §102 bite.
2. Google Patents "Patent Citations" tab for US8556983B2 (https://patents.google.com/patent/US8556983B2/en) — export the full "Cited By"/"Citations" lists; I could not reach the citation table in the fetched text.
3. FreePatentsOnline (https://www.freepatentsonline.com/8556983.html) — its front-page reproduction includes the "(56)" block; use it to cross-verify the numbers I listed and to get the entries above 5,871,542 that my excerpt missed.
4. PTAB / IPR record — IPR2017‑00115 and IPR2017‑00779 exhibit lists give you the Petitioners' art and the experts' claim-element mappings, which are the best available proxy for how this family's claims were actually attacked.
5. Effective-date determination — compare the granted claim language to the 12/660,529 and 61/339,766 disclosures under §112 to fix the priority date. Without that step, a per-reference §102 table is unreliable.

Honest statement of limits

I have not verified the complete front-page reference list for US 8,556,983, not examined its file wrapper, and not confirmed that the reference window I captured is co-extensive with the actual "(56)" block. Where I gave a "brief description," it is my characterisation from general knowledge or from the '983 specification's own discussion of the reference; entries marked unverified are exactly that. Given the STRICT RULE, I have preserved every identifier literally as retrieved (including 6,013,103 Kaufman and 6,197,325 MacPhee) rather than substituting numbers I believe "should" be there. Against the granted claim set as characterised earlier, my screen finds no §102 anticipation from the verified cited art, with the strongest (though still incomplete) candidates being the patient-specific CAD/CAM and cartilage-measurement references — D'Urso 6,112,109, Ateshian 6,126,690, Aouni‑Ateshian 6,161,080, Votruba 5,899,859, and DiGioia 5,995,738/6,002,859 — none of which reaches the minimum/maximum implant-thickness limitation that defines independents 1, 9, and 16.

I'll research the prior art references cited on this patent and in the surrounding ConforMIS family before building the § 103 analysis.

Let me pull the full "References Cited" list and the key prior-art references relevant to thickness-constrained, patient-matched implant design.

Let me gather the cited references and the key FEA/patient-specific design prior art.

Obviousness Analysis — US 8,556,983 B2 under 35 U.S.C. § 103

Scope note / limitation on sourcing. The live page's "Prior Art" section as provided in this session is truncated and does not include the patent's full "References Cited" list. The analysis below therefore rests on (i) the third‑party references I could positively confirm from the public record for this patent and its family (Google Patents, FreePatentsOnline, uspto.report, Justia, and patent‑office search reports citing the '983 family), and (ii) the background art the specification itself identifies. Where I could not verify a citation number or its precise disclosure, I flag it. Do not treat any reference I could not retrieve as absent from the patent's face.

1. The claims to be tested

This analysis takes as authoritative the grant claim set identified in the prior section: 18 claims, independents 1, 9, 16, all method claims (that section already flags that the count should be re‑verified against the official printed patent).

Dependents add: implant type (2, 3), CAD/CAM (4), altering a blank (5), curve smoothing (6), FEA (7, 12, 14), load‑bearing analysis (8, 13, 15), increasing radius of curvature (10, 17), and dual min/max thresholds (11, 18).

Claim‑construction observations that drive the obviousness case. "Substantially matches … or a predetermined percentage of" (claim 1(b)) is a broadened, non‑exact match — the patentee disclaimed exactness on the face of the claim. "Selecting and/or designing" is likewise disjunctive. These constructions make the claims easier, not harder, to read onto the art.

2. Level of ordinary skill (POSITA)

A person of ordinary skill would be a biomechanical/orthopedic design engineer, or an orthopedic surgeon‑engineer, with a working knowledge of (i) CT/MRI‑based 3‑D modeling of joints, (ii) CAD/CAM implant manufacture, (iii) knee arthroplasty resection geometry, and (iv) finite‑element/load analysis of load‑bearing implants. Every one of those four competencies was independently established in the art by the 2001‑2007 window, as the family's own citation list confirms (e.g., Delp et al., Comput. Biol. Med. 25(1):21‑34 (1995) graphics‑based musculoskeletal modeling; Thoma et al., Der Orthopäde 29(7):641‑644 (2000), custom knee endoprostheses from 3‑D CT; Harrysson et al., FEA of custom femoral components from patient CT).

3. The prior‑art references

4. The obviousness case

4.1 Claim 1 — Fell alone, or Fell + White + Ries

Fell maps with unusual completeness:

• (a) → Fell's MRI‑derived "radius of curvature of the femoral condyle R_FC" and tibial plateau radius: a curvature on an articular surface identified preoperatively from patient‑specific data.
• (b) → Fell's outer contours "designed to substantially mate with the corresponding tibial and femoral surfaces" / "the symmetry of the device's femoral side may be matched with the femoral condyle."
• (c) → "a determination is made of the current joint spacing … This measurement determines the thickness of the device," plus Fell's express thickness ranges.
• (d) → Fell's Tibial Base Plane is a mathematically defined planar inner facet, and the device thickness is by construction the distance from that plane to the upper bearing surface. Fell also teaches biasing the thickness distribution across the device.

If the patentee distinguishes on the ground that Fell's device is free‑floating (no resection facets), White and Ries close that gap: White teaches the same scan→model→customized‑implant workflow with a standard (e.g., thickness) feature, and Ries/standard arthroplasty practice teaches that a resurfacing component's inner surface is planar cut facets. A POSITA placing Fell's thickness‑controlled geometry onto a faceted resurfacing component is doing nothing more than adapting a known design to a known use.

4.2 Claims 9 and 16 — the curvature‑modification steps

These are the only steps not literally disclosed by Fell, and the obviousness case is a classic KSR design‑constraint argument:

1. Fell already teaches departing from the anatomical curvature. Fell expressly says "it is desirable to offset the contour angles (from the CSO) … This is done to bias the thickness of the meniscal device to the periphery," and that devices may need "considerably greater thickness" for deformed/worn anatomy. That is modifying a patient‑derived curvature to achieve a target thickness distribution — the same operation as claims 9(c) and 16(c), differing only in that the '983 names the constraint as a threshold.
2. The two constraints are the two universal, opposed design objectives of the field. A resurfacing implant must simultaneously be (i) thick enough not to fracture/wear (a minimum thickness floor — standard metal and polyethylene design practice, and expressly present as a range in Fell) and (ii) thin enough to preserve bone (a maximum thickness ceiling — the entire stated purpose of Radermacher/Thoma custom endoprostheses and the '983's own "pre‑primary" framing).
3. Motivation to combine. (a) Design incentive: the specification itself concedes the field's goal was minimizing resected bone; a POSITA optimizing for bone preservation must cap thickness. (b) Predictable result: changing a radius of curvature by a small amount to satisfy a thin/thick constraint is a monotonic, predictable geometric adjustment; claim 1's own "predetermined percentage" language admits the match need not be exact. (c) Finite, identified solutions: for a given patient curvature and a given threshold, the set of compliant curvatures is a small, bounded family (claims 10 and 17 even name the only direction for the minimum‑thickness case — "increasing the radius of curvature," a wholly conventional move that flattens a condyle).

4.3 Claims 7, 8, 12, 13, 14, 15 — FEA and load‑bearing analysis

These dependents are, standing alone, highly vulnerable. Harrysson (2007) teaches using FEA on patient‑specific CT data to evaluate a custom femoral component; FEA and load‑bearing analysis were routine orthopedic design verification tools by the relevant date (Delp 1995; the family's own citation list). Because claim 9/16 already recite enforcing a thickness threshold — an inherently structural justification — using FEA or load analysis to select or confirm that threshold would have been an obvious verification step ("applying a known technique … to yield predictable results"). These dependents rise and fall with their independents; if 1/9/16 fall, they add little.

4.4 Claims 10 and 17 — "increasing the radius of curvature"

Claim 10 (min‑thickness) and 17 both recite merely increasing the radius. Fell computes and uses radii of curvature throughout; Kreiss/standard femoral sagittal J‑curve design teaches variable/larger radii at the condyle. Increasing a radius to thin a component is the single most obvious mechanical step available and is essentially an instruction to use the only direction that reduces thickness.

5. KSR rationales to be pleaded

6. Strongest patentee counterarguments (and why they are likely to fail)

1. "Different field of endeavor — Fell is a meniscal spacer, not a resurfacing arthroplasty implant." Weak. Both are patient‑imaged, patient‑shaped articular knee implants addressing joint congruity, load distribution, and bone preservation; the '983 Background itself cites the same arthroplasty literature.
2. "All references teach matching the patient's curvature; none teach modifying it." Fell affirmatively contradicts this — its contour‑angle offset to "bias the thickness" is a modification of the patient‑derived geometry to serve a design constraint. That undercuts any teaching‑away argument.
3. "The prior art does not disclose a min/max thickness threshold." Fell discloses numerical thickness ranges and individual tailoring; standard metal/polyethylene minimum thicknesses were ordinary design knowledge.
4. Secondary considerations. None of record in this session. If ConforMIS held objective‑indicia evidence (commercial success of patient‑specific implants, praise, licensing), that would be the strongest (indeed the only) meaningful rebuttal; I found none in the sources I could access, but I did not search litigation/exhibit databases directly.

7. Bottom line

• Claim 1: likely obvious over Fell '927/'421 alone, and clearly obvious over Fell + White + Ries. Fell supplies every step but the explicit "facet" vocabulary.
• Claims 9 and 16: obvious over Fell + Radermacher/Thoma, with the min/max thickness thresholds supplied by ordinary design knowledge and admitted by the specification; the "modify the curvature" step is the KSR design‑constraint case. Fell's express "bias the thickness" teaching is the single most damaging piece of art to these two independents.
• Claims 7, 8, 12, 13, 14, 15: obvious over Harrysson (2007) and routine FEA/load‑analysis practice.
• Claims 10/17: obvious as a strictly mechanical radius increase once a minimum thickness is the goal.

Practical caveat: because the patent shows Expired – Lifetime (anticipated expiration 2022‑05‑28), any § 103 challenge today is only worth mounting to defeat past damages/ongoing royalties or in a defensive posture; an IPR remains available against an expired patent for claim cancellation but not amendment.

Verification flags for the record:

1. I could not retrieve the patent's full "References Cited" list in this session; the Fell and White references are confirmed from public sources, but US 5,768,134 (Radermacher) and the precise '983 face citations should be verified against the printed patent / PatentCenter.
2. Claim count, independents, and exact dependent mapping are taken from the earlier section, which itself flagged them as unverified.
3. The Harrysson and Thoma citations are confirmed as appearing in the ConforMIS family's reference lists, not necessarily on the '983 face.
4. No secondary‑considerations evidence and no prosecution‑history statement of allowability was located; both could materially change the analysis.