US 11883219

I'll search for this specific patent number and related litigation dockets.

Let me pull the actual claim text and check for any 2026 litigation activity.

US 11,883,219 B2 — Analyst Summary

Scope note before the substance: I was able to confirm the bibliographic record from multiple independent sources (Google Patents, Justia, and third-party mirrors all agree). However, no litigation — district court or CAFC — was located for this patent number, and the authoritative full-text I retrieved was truncated before the claims section. I flag that limitation explicitly below rather than paper over it.

1. Bibliographic Data

Chain of title (per recorded assignments): OrthoGrid Systems, Inc. → OrthoGrid Systems Holdings, LLC (assigned Aug. 26, 2022) → OrthoGrid Systems, Inc. (assigned Sept. 23, 2024). Separately, Zimmer Biomet Holdings announced on Aug. 7, 2024 a definitive agreement to acquire OrthoGrid Systems, Inc., expressly citing the AI-driven Hip AI® platform and "over 40 patents" — so the patent is plausibly now beneficially held within Zimmer Biomet, though I did not find a recorded assignment of this specific patent to Zimmer Biomet, and the acquisition was announced as subject to closing conditions.

2. Abstract (verbatim substance)

The abstract claims a computing platform executing one or more automated AI models, including a neural network model, trained on a plurality of radiographic images from a data layer to detect a plurality of anatomical structures or a plurality of hardware — wherein at least one anatomical structure is a pelvic teardrop and a symphysis pubis joint; detecting those structures in a subject radiographic image by classifying the image with reference to a subject good-side radiographic image; and constructing a graphical representation, namely a subject-specific functional pelvis grid generated from the detected anatomical structures.

3. Independent Claims — Plain-Language Overview

⚠️ Confidence caveat: The full text I retrieved cuts off mid-specification (at the intra-operative reaming/functional-pelvis-grid discussion), so I do not have 11,883,219's own claim set in front of me. The abstract of '219 is word-for-word identical to claim 1 of US 11,540,794, a same-family OrthoGrid patent I did retrieve. I therefore reconstruct the independent claims from that parallel family member and the '219 abstract, and I mark this as inference, not verified claim text.

Independent Claim 1 — Method of intra-operative surgical imaging (method claim):

1. Provide a computing platform with at least one image-processing algorithm for classifying radiographic orthopedic images, the platform executing automated AI models (including a neural network) trained on a data layer of radiographic images to detect anatomical structures and/or hardware — the anatomical structure being a pelvic teardrop and/or a symphysis pubis joint;
2. Detect those anatomical structures in a subject's radiographic image, where detection is performed by classifying the radiographic image with reference to a subject good-side (contralateral) radiographic image; and
3. Construct a graphical representation — a subject-specific functional pelvis grid — generated from the detected structures.

Gist: AI landmark detection on a hip/pelvis radiograph, benchmarked against the patient's unaffected side, producing a patient-specific reference grid overlay — intended for intra-operative fluoroscopy in total hip arthroplasty rather than CT-based navigation.

Independent Claim 9 — System claim:
An AI-based intra-operative surgical guidance system comprising a non-transitory computer-readable medium encoded with instructions forming a software module with three layers — a data layer, an algorithm layer, and an application layer — plus a processor. The system is trained to detect anatomical structures, detects them in a subject radiograph by classifying against a subject good-side image, and constructs a subject-specific functional pelvis grid.

Notable dependent claims (from the parallel family member):

• Registering the good-side image to the operative-side image (cl. 2)
• Registering pre-op ipsilateral to post-op ipsilateral images (cl. 3)
• Constructing an anatomy map of an entire anatomical region (cl. 4)
• Quantifying measurements between structures, receiving a required output from a user, and calculating it from vector or vertical distances (cl. 5)
• Output = leg length determination (cl. 6) or offset determination (cl. 7)
• Detecting hardware in the radiograph by classifying against the good-side image (cl. 8)
• Anatomy map generated from detected structures in the system claim (cl. 10)

4. Specification Highlights (for context)

The disclosure describes a layered architecture (data layer 105 → algorithm layer/"AI engine" 106 → application layer 107), CNNs for optimal-vs-suboptimal outcome classification, reinforcement learning for screw/instrument trajectory, distortion-corrected dimensioned grids, "ghosting" (overlay of good-side on affected-side images with a match-confidence percentage), statistical shape models for 3D-from-2D inference, and a family of functional grids: functional pelvis grid, neck-cut grid, reamer-depth grid, center-of-rotation grid, and cup-abduction grid. Non-image clinical text may be processed via NLP, and a Cox proportional-hazards or similar regression model may serve as an alternative outcomes classifier.

5. Litigation / CAFC Status

No results. My searches for this patent number in litigation contexts — including targeted queries against Federal Circuit and district-court material, and OrthoGrid/Zimmer Biomet patent disputes — returned no case, docket entry, IPR, or appeal referencing US 11,883,219. The hits that surfaced were unrelated Rule 19 / sovereign-immunity and venue disputes that merely happened to contain the search terms.

Because the tooling available to me is general web search rather than a direct PACER, Docket Navigator, or USPTO Patent Trial and Appeal Board API query, I cannot certify that no 2026 CAFC docket exists — only that none is discoverable through the searches performed. If you need an authoritative negative, that requires a direct PACER/PTAB database pull.

6. Explicit Uncertainties

1. Claim text of '219 itself is reconstructed, not read. Based on the identical abstract/claim-1 language in same-family US 11,540,794. Verify against the granted '219 claim set before relying on it.
2. Current ownership is ambiguous. Google Patents and the recorded assignments point to OrthoGrid Systems, Inc.; Justia still lists OrthoGrid Systems Holdings, LLC. The Zimmer Biomet acquisition was announced, and I found no instrument transferring this patent to Zimmer Biomet.
3. The full-text retrieval was truncated mid-specification, so I cannot characterize the drawing-by-drawing enablement or the prosecution history (e.g., what was amended over the cited PCT search report art).
4. No legal-status analysis performed. "Active" reflects the source's label; no maintenance-fee or reexamination/reexam-certificate check was run.

I'll search for litigation involving US patent 11883219.

Let me search more targeted litigation databases.

Litigation Search: U.S. Patent No. 11,883,219

Bottom Line

I found no litigation that specifically names U.S. Patent No. 11,883,219 as an asserted or challenged patent. Based on the searches run (general web, court-opinion repositories surfaced via CourtListener, PTAB/USPTO document repositories, and Google Patents' litigation links), I could not identify a plaintiff, defendant, case number, or filing date for any case asserting or challenging this particular patent. I want to be explicit that this is a negative finding, not a certification — see the caveats below.

What the searches did return (relevant context)

Why my answer is hedged

• No live access to the sources you named. I could not directly query the Unified Patents litigation portal (portal.unifiedpatents.com/litigation/caselist), PACER (paywalled), or Docket Navigator / Lex Machina. My searches were web searches, and they returned no docket record naming '219.
• Google Patents' "litigation" linkage for this patent produced no case entries in the fetched text.
• The ambiguous E.D. Tex. / IPR2025-01027 lead is the one item worth chasing down, and I could not confirm the patent number. The consolidation date (Feb 18, 2025) post-dates the Zimmer Biomet acquisition announcement, so it is plausible but unverified that it relates to the OrthoGrid portfolio.

Recommended verification steps (for the file)

To definitively confirm or exclude litigation, run these before relying on the negative finding:

1. PACER / CourtListener RECAP — query party name "OrthoGrid Systems" (and "Zimmer Biomet" as co-plaintiff) across all districts; then check the "patents" field of any hit for 11,883,219.
2. USPTO PTAB (PTACTS) — pull IPR2025-01027 and read the challenged patent number; also check for any IPR/PGR naming 11,883,219 directly.
3. Unified Patents litigation caselist — search "OrthoGrid" and "Zimmer Biomet" as patent owner.
4. USPTO Assignment records — confirm current ownership/standing, since the '219 patent (application 17/895,948) underwent two recorded reassignments (2022-08-26 to OrthoGrid Systems Holdings LLC; 2024-09-23 back to OrthoGrid Systems, Inc.).

Caveat on identifiers

Per your instruction, I have interpreted the number literally as 11,883,219. All statements above refer exclusively to that patent (application 17/895,948, "Artificial intelligence intra-operative surgical guidance system and method of use"). I have not substituted or normalized any similar number. If no case number was found for this patent, I have said so rather than reporting a case from a neighboring patent number.

Status: No confirmed litigation identified as of the search date.

I'll verify the structured block's "no proceedings" default against live sources before writing.

Initial searches surfaced nothing on point. Let me run targeted queries on the patent owner and its litigation history.

Let me check dockets and any PTAB filings tied to the family's later patents.

Proceedings overview

Total AIA trial proceedings on file for US 11,883,219: 0 (0 active, 0 claims invalidated, 0 claims sustained, 0 settled, 0 institution denials). The USPTO ODP structured block reports no AIA trial proceedings, and my independent web searches — for the patent number, the patent owner (OrthoGrid Systems / OrthoGrid Systems Holdings), and the asserted-art docket sources — surfaced no IPR, PGR, or CBM naming this patent or this patent owner. Bottom line for a defendant: there is no PTAB record to lean on. The patent has not been tested at the Board, so you cannot point to a canceled claim, an estoppel-barred petitioner, or a Board finding of unpatentability. The entire invalidity case — including any § 102/§ 103 art and any § 112 or § 101 theory that would have been teed up in a PGR — remains unadjudicated and, critically, un-estopped for whoever moves first.

Because the count is zero, the per-proceeding template below is empty by construction. I have not substituted guesses or filled it with proceedings from other patents in the family.

(No proceedings to itemize)

No IPR, PGR, or CBM proceeding number exists to report. I am declining to populate the proceeding sections with anything, per the constraint against inventing proceeding numbers.

One false-positive worth flagging so your team does not repeat my search path: searching the numeric neighborhood returns IPR2025-00249, which involves U.S. Patent No. 11,318,227 — a collagen-fiber/collector patent, not OrthoGrid's 11,883,219. The digit strings are close; the subject matter is unrelated. Do not let a keyword hit on "IPR2025-00249" migrate into a memo as pertaining to the OrthoGrid patent. (IPR2025-00249 sur-reply, PTAB)

Strategic summary

Claim status: every claim is UNTESTED. With zero PTAB proceedings, there is no canceled-claim set, no surviving-claim set, and no Board-construed claim scope. All claims of 11,883,219 — including the independent claims directed to the computing platform with a neural-network model trained to detect anatomical structures (a pelvic teardrop and symphysis pubis joint) and to construct a subject-specific functional pelvis grid, and the independent claims to the AI-assisted total hip arthroplasty method and the software-module system (data layer / algorithm layer / application layer) — stand exactly as issued on 2024-01-30. If a demand letter or complaint asserts this patent, no claim has been weakened by an adverse PTAB judgment, and conversely no claim has been strengthened by a PTAB win. Note also that the specification describes a large claim family around the same disclosure (related grants in the portfolio include U.S. 10,973,590, 11,386,556, and 11,937,888, per public portfolio listings) — an adverse result on 11,883,219 alone would not necessarily clear the portfolio.

Estoppel landscape: no § 315(e)(2) estoppel attaches to anyone. Estoppel runs from a petitioner's IPR/PGR that reaches a final written decision. With no petition and no FWD, there are no barred petitioners, no privity chains to trace, and no "reasonably could have raised" carve-outs to map. For a defendant being asserted against today, that is a structural advantage on the invalidity side: you can raise any prior-art ground, in any forum, without worrying that a predecessor petitioner burned it. It also means there is no free ride — nobody else has done the work, developed the expert record, or produced a Board-tested invalidity theory you can borrow. If you want PTAB relief, you (or a proxy) have to file it and fund it.

Pattern signals: none available. No petitioner has filed even once, so there is no serial-filer pattern, no joinder history, and no indication of a defensive aggregator (e.g., Unified Patents) in the chain. I found no evidence that the patent owner has pursued PTAB appeals — there is no FWD to appeal. On ownership: the structured record shows current assignee OrthoGrid Systems, Inc., with an assignment recorded 2024-09-23 out of OrthoGrid Systems Holdings, LLC; separately, Zimmer Biomet announced a definitive agreement to acquire OrthoGrid Systems, Inc. on 2024-08-07, explicitly citing "over 40 patents" in the deal (Zimmer Biomet press release). For your risk model, the practical read is that the patent now sits inside a large strategic acquirer's portfolio rather than a small non-practicing entity's hands — which tends to change assertion dynamics (more likely defensive/competitive cross-licensing pressure, less likely demand-letter monetization) even though it does nothing to the legal status of the claims. I could not confirm from public sources any district-court or ITC action actually asserting 11,883,219; treat that as unverified rather than as proof no suit exists.

Timing/vehicle note for your team: the PGR window is closed (9 months from the 2024-01-30 grant ran out around 2024-10-30), and CBM is unavailable (the transitional program sunset in 2020 and, in any event, these are orthopedic-imaging claims, not financial-services claims). IPR under § 311 is the only live AIA vehicle — and it is unlimited in time after grant unless you have been served with a complaint alleging infringement, in which case § 315(b)'s one-year bar applies to you and runs from service.

Recommended next steps

1. Verify the zero directly, and read the absence as a real data point, not an ingest gap. Pull the patent's "Proceedings" tab on the USPTO PTAB E2E / PTAB Center and the patent-specific page on Patent Center for 11,883,219. Then cross-check the owner's litigation posture via the PTAB E2E search filtered on "OrthoGrid," plus CourtListener and the Federal Circuit docket for any appeal involving OrthoGrid or Zimmer Biomet. If it confirms clean, cite that confirmation in your file. The absence of PTAB activity is itself a signal: heavily asserted patents almost invariably attract IPRs, so a clean record hints that this patent may not yet be in active monetization — which cuts against a "the asserted claims have already been killed" defense and in favor of negotiating from a clean slate.

2. If you are a defendant served with a complaint citing 11,883,219: calendar the § 315(b) one-year bar from the service date immediately — that is your only hard IPR deadline. If you file, the § 316(a)(11) clock gives the Board one year from institution to reach a final written decision, and institution itself is decided by statute within six months of the petition's filing. Build the petition against the issued independent claims and have your expert address the fluoroscopy-distortion, landmark-detection, and grid-registration limitations head-on, since the specification affirmatively relies on those features for novelty (see, e.g., the § 112-style enablement/priority discussion around the priority chain to PCT/US2019/050745, published as WO2020056086A1, which is the family member you should mine for written-description and priority challenges).

3. If you are a defendant and no proceeding exists (your current posture): do not count on PTAB help that was never obtained. Your invalidity case must be built from scratch — prior art search, expert declaration, claim construction — and you should weigh filing your own IPR as a stay-and-leverage play versus preserving grounds for district court (filing an IPR triggers § 315(e)(2) estoppel against you on grounds you raised or reasonably could have raised, so decide deliberately). Also confirm the current owner's identity before responding to any demand: the 2024-09-23 assignment record and the Zimmer Biomet acquisition announcement both suggest the address of record may have changed, and you want your correspondence going to the real party in interest.

Confidence and limitations: I have high confidence that no PTAB proceeding is currently indexed for this patent, based on the structured ODP block you provided plus my searches. I have low confidence that this reflects a permanently clean record — ODP ingest lags by days-to-weeks, and a recently filed petition might not yet appear. I expressly do not know of any FWD, institution decision, settlement, or Federal Circuit appeal for this patent, and I found none; if a proceeding exists outside the indexed sources, the claim-level outcomes, panel composition, and disposition are unknown to me and should be confirmed at PTAB E2E before you rely on the "zero" in any filing or client memo.

I'll research the assignment chain for US 11,883,219. Let me run several parallel searches.

Let me dig into the specific assignment records and inventor details.

Assignment Reconstruction: U.S. Patent No. 11,883,219

Sourcing caution up front: I was able to reach Google Patents' legal-events timeline and secondary sources, but I could not get a live read of the USPTO Patent Assignment Center record itself (assignmentcenter.uspto.gov / assignment.uspto.gov), and no source I retrieved exposed the reel/frame numbers for this patent. Dates below are grounded in Google Patents legal events and corroborating press/8-K coverage; reel/frame entries are marked as unverified rather than invented. I did not fabricate any reel/frame. Treat the reel/frame fields as a to-do, not a finding.

Inventors

Named inventors on the face of the '219 patent (also reflected in the related family member US 11,540,794, "Boddington et al."):

Patterns: No departure pattern precedes the transaction. Both named founders (Boddington, Saget) were still serving as co-CEOs and were quoted in the acquirer's announcement, i.e., they stayed through the exit. Oulhaj's Strasbourg address aligns with OrthoGrid's French R&D site — this is a distributed operating-company team, not a rented-inventor arrangement. Kubiak is a recurring surgeon-inventor across the OrthoGrid grid portfolio, consistent with a university/clinician collaboration rather than a standalone assignment.

Original assignee

• Entity on the issued patent: Google Patents lists the Original Assignee as ORTHOGRID SYSTEMS HOLDINGS, LLC, with the Current Assignee as ORTHOGRID SYSTEMS, INC.
• Primary line of business: OrthoGrid Systems, Inc. was an operating medical-device/software company (founded 2012, Salt Lake City, UT) selling Hip AI®, an FDA-cleared, fluoroscopy-based AI surgical guidance platform for total hip arthroplasty, plus two other FDA-cleared applications. Product embodiment is documented: hospital license/software agreements (e.g., University Medical Center of Southern Nevada) and a Hip AI commercial launch ahead of the acquisition.
• Status: Acquired by Zimmer Biomet Holdings, Inc. (NYSE/SIX: ZBH). Definitive agreement announced 2024-08-07, "expected to be completed by the end of the fourth quarter of 2024"; Zimmer Biomet's 2024-10-28 release states the acquisition was completed "earlier this month" (i.e., October 2024). The deal explicitly included "over 40 patents." OrthoGrid is therefore now a Zimmer Biomet subsidiary; the LLC named as original assignee is best understood as the company's IP-holding affiliate, not an independent shell.

Assignment timeline

Two post-filing ownership events are corroborated by Google Patents legal events. Reel/frame numbers could not be verified and are deliberately omitted rather than guessed.

• Executed ~2022-08-25 / recorded 2022-08-26 — Reel UNVERIFIED

  • Conveyance: Assignment (recorded as "ASSIGNMENT OF ASSIGNORS INTEREST")
  • Assignor: OrthoGrid Systems, Inc.
  • Assignee: OrthoGrid Systems Holdings, LLC
  • Correspondent: Not retrievable (assignment-center record not accessed). Flag: the OrthoGrid portfolio's prosecution counsel of record has recurred across family patents as a single attorney — "Susan B." first at Harris, Shelton, Hanover & Walsh (US 9,456,874) and later at Veritay Group IP, PLLC (US 11,386,556). This is a prosecution counselor, not confirmed as the assignment-record correspondent. See Signal 3.
  • Context: Internal IP consolidation — operating company's patent estate moved into its "Holdings" affiliate the day after the '219 application was filed (App. 17/895,948, filed 2022-08-25).

• Executed ~2024-09 / recorded 2024-09-23 — Reel UNVERIFIED

  • Conveyance: Assignment (recorded as "ASSIGNMENT OF ASSIGNORS INTEREST")
  • Assignor: OrthoGrid Systems Holdings, LLC
  • Assignee: OrthoGrid Systems, Inc.
  • Correspondent: Not retrievable.
  • Context: M&A/title cleanup — IP moved back from the holding affiliate into the operating entity approximately six weeks after the Zimmer Biomet deal was announced (Aug 7, 2024) and just before the anticipated Q4 2024 close, positioning clean title in the acquisition target.

• No recorded assignment to Zimmer Biomet appears in the legal-events timeline as fetched (fetched 2026-09-24). The likely explanation is a stock/equity purchase — Zimmer Biomet acquired OrthoGrid Systems, Inc. itself, so patent title stayed with OrthoGrid Systems, Inc. as a ZB subsidiary and no title-transfer record was required. This is inference, not a confirmed record; verify via the Assignment Center.

Other legal events (non-assignment, for context): filed 2022-08-25 (App. 17/895,948); published 2023-01-05 as US20230000451A1; granted 2024-01-30; anticipated expiration 2039-09-12. Priority date 2018-09-12, claimed from PCT/US2019/050745.

Timeline diagram

timeline
    title Ownership of US 11883219
    2012 : OrthoGrid Systems founded
    2018 : Priority date from PCT filing
    2022 : Application filed 25 Aug
         : Assigned to OrthoGrid Systems Holdings LLC
    2023 : Published as US20230000451A1
    2024 : Patent granted 30 Jan
         : Assigned back to OrthoGrid Systems Inc
         : Zimmer Biomet announces acquisition
         : Acquirer closes deal in October

NPE / troll-pattern signals

Verdict

Operating-company assertion — with an explicit caveat: the ownership character is unambiguously that of a product-shipping operating company (OrthoGrid → Zimmer Biomet, NYSE: ZBH), and none of the eight NPE signals is present on confirmed evidence. The chain is short, internal, and M&A-driven (Holdings transfer 2022-08-26; transfer back to the operating entity recorded 2024-09-23, immediately bracketing the Zimmer Biomet acquisition announced 2024-08-07 and closed Oct 2024). Important caveat: the taxonomy's "assertion" prong is unproven here — no litigation asserting '219 has been identified (see the earlier litigation section), so the most precise statement is "operating-company ownership, currently non-asserting." I selected the operating-company verdict rather than "Defensive/non-asserting" because the chain does not terminate at a defensive aggregator (Signal 8), which is the defining feature of that category.

Verification links:

• USPTO Patent Assignment Center: https://assignmentcenter.uspto.gov/ (search patent number 11883219)
• Legacy assignment search: https://assignment.uspto.gov/patent/index.html
• Google Patents record: https://patents.google.com/patent/US11883219/en

Open items to close the file: (1) pull the two assignment records to capture reel/frame and the assignment correspondent; (2) confirm whether title to '219 transferred to Zimmer Biomet or remains with OrthoGrid Systems, Inc. as a ZB subsidiary; (3) reconcile against the previously flagged (and still unconfirmed) E.D. Tex. / IPR2025-01027 lead.

Identifier note: all statements refer strictly to US 11,883,219 (App. 17/895,948). Where the same record could not be confirmed, I said so rather than borrowing data from family members such as US 11,540,794 or US 11,386,556, which are cited only as corroborating context for inventor identities and prosecution counsel.

I'll search for the patent's cited references and prior art. Let me run several targeted searches.

My searches returned generic results rather than the specific citation list. Let me target the patent text directly.

Prior-Art Analysis: U.S. Patent No. 11,883,219 B2

Important up-front limitation. The task asks me to "look at each patent citation for 11883219" and map each to claims under § 102. I have to be transparent: my searches did not return the front-page "References Cited" list for U.S. 11,883,219 B2, and the full patent text you supplied is truncated before that section (it ends mid-sentence in the reaming discussion, with no "References Cited," no claims, and no bibliographic data). I cannot enumerate the citation list from memory without risking fabrication, and the third-party databases that would carry it (USPTO PatentCenter "References Cited," Google Patents "Patent Citations") did not surface in the results returned.

So this section is organized as: (A) what is confirmed, (B) what is a candidate with an explicit caveat, (C) what is expressly excluded as non-prior art, and (D) a precise completion procedure. I flag every inference rather than presenting it as fact.

A. Confirmed bibliographic anchor for '219 (treat literally per your instruction)

B. Reference confirmed inside the '219 specification (highest confidence)

U.S. Pat. No. 9,610,134 — expressly incorporated by reference in the specification:

"…the identification of corresponding specific anatomical structures in preoperative 115 and intraoperative images 120. See, e.g., U.S. Pat. No. 9,610,134 specifically incorporated by reference in its entirety."

• Status: Named in the patent itself → confirmed to exist as a reference in the file, but this is an incorporation by reference for the image-mapping/registration method, not a listed front-page citation.
• Date: I could not confirm its issue date or title within the search budget — do not treat any date I might infer as verified.
• § 102/§ 103 relevance: Because the mapping/registration teaching is incorporated, the '219 claims that recite "computing a best-fit image transformation from the preoperative to the intraoperative image space" / registration of a dimensioned grid to anatomy (the anatomy-map and grid-registration subject matter in the Definitions section) would need to be read against this reference. If it pre-dates 2018-09-12 it is § 102(a)(2)/102(b) art at minimum for that feature; where only a sub-element is disclosed, it is more naturally a § 103 combination reference. I cannot responsibly state which specific claim numbers it anticipates without the reference and the '219 claim set in hand.

C. Candidate third-party patent references (citation relationship observed — direction unconfirmed)

The following appeared in citation tables on Google Patents with U.S. 11,883,219 B2 listed among the citing/cited documents. The listing format (publication number / priority date / publication date / assignee / title) is Google Patents' table format, which is consistent with '219 appearing in the reference's "Cited By" relationship — i.e., these are candidate references cited by the '219 family, not later art citing '219:

• Caveat (must be preserved in the file): I am inferring the citation direction from the table layout in search snippets. Google Patents also renders a "Citations" (outgoing) table in the same visual format. I could not confirm whether these are outgoing citations of '219 or incoming citations to it. Do not rely on these as prior art without verifying the direction and the dates in PatentCenter.
• Even if they are outgoing citations, US 2011/0152676 and US 2015/0086955 both pre-date 2018-09-12 on their faces, so either could be § 102(a)(1)/102(a)(2) art if the direction is confirmed.

D. Documents expressly NOT prior art to '219 (to avoid a common analysis error)

These surfaced in my searches but are same-family / commonly-owned disclosures and therefore are not, or are excepted as, prior art:

E. § 102 claim mapping — what I can and cannot state

The Definitions section you supplied identifies the independent-claim themes of '219:

1. AI system: computing platform executing models trained on a data layer (at least surgical images) → calculate intra-operative surgical decision risks → provide guidance → visual display → classification algorithm.
2. Neural-network system trained on radiographic images to detect anatomical structures (pelvic teardrop, symphysis pubis) or hardware, classifying the radiograph with reference to a subject's good-side image, and constructing a subject-specific functional pelvis grid.
3. AI-assisted THA method: preoperative radiograph → detect anatomical structures → generate subject-specific functional pelvis grid → receive intra-op AP pelvis radiograph → identify anatomical landmark/hardware → guidance.
4. AI system with data layer + algorithm layer + application layer and the good-side-referenced pelvis-grid construction.
5. Anatomy-map method: construct anatomy map of an entire anatomical region → quantify measurements between structures → receive required output from user → calculate based on vector/vertical distances.

Honest statement: I cannot assign any of these specific claims to a specific reference under § 102 with the evidence currently in hand, because (i) I do not have the '219 front-page citation list, and (ii) I do not have the issued claim text. Doing so would require me to invent a reference-to-claim mapping, which I will not do.

The most likely § 102 battleground, if you recover the reference list, is the generic AI-guided registration/classification subject matter (theme 1 and the registration modules) — the US 2011/0152676 and US 2015/0086955 candidates live in exactly that space. Conversely, the subject-specific functional pelvis grid keyed to teardrop + symphysis pubis landmark detection classified against the good-side image (themes 2–4) is comparatively narrow and is where the OrthoGrid family itself is most densely patented — meaning the strongest § 102 challenges there will most likely be the applicant's own earlier applications rather than third-party art, and those may be excepted under § 102(b)(2)(C) only if the § 102(a)(2) route is used.

F. To complete this properly (recommended steps for the file)

1. USPTO PatentCenter / Patent Application 17/895,948 → "References Cited" — this yields the authoritative front-page list (U.S. patent documents, foreign documents, and non-patent literature) for U.S. 11,883,219.
2. Google Patents → US11883219B2 → "Patent Citations," "Cited By," and "Similar Documents" — pull the three tables separately, since direction is the trap noted in Section C.
3. Pull U.S. Pat. No. 9,610,134 and confirm its issue date, title, and assignee, then identify which interior paragraphs supply the mapping/registration teaching the '219 spec relies on.
4. Confirm the '219 issued claims (count and text) from PatentCenter — I could not confirm the claim count for '219 (the sibling U.S. 11,540,794 has 11 claims and 63 drawing sheets, which is not evidence about '219's claim set — do not import it).
5. Check the IDS history in the file wrapper for references cited during prosecution that are not on the printed front page.

G. Explicit uncertainty list (what I am not asserting)

• I am not asserting that US 2011/0152676, US 2015/0086955, or US 2022/0233159 are (or are not) prior art to '219 — direction and effective dates are unverified.
• I am not asserting any date, title, or assignee for U.S. Pat. No. 9,610,134.
• I am not asserting a claim-count for '219.
• I am not asserting the absence of other cited references — the list was not retrievable, and a negative finding from my searches is not a certification of completeness.
• Consistent with the prior litigation section, no issuance/ownership facts above contradict that section; the two agree that OrthoGrid Systems, Inc. is the current owner of record (2024-09-23).

Bottom line: only one reference — U.S. Pat. No. 9,610,134, expressly incorporated in the '219 specification — is confirmed. Two third-party candidate references (US 2011/0152676 A1, US 2015/0086955 A1) are identified as probable cited art but require direction/date verification. No defensible § 102 claim-by-claim anticipation mapping can be produced until the front-page "References Cited" list and the issued '219 claims are retrieved from USPTO PatentCenter.

I'll research the prior art landscape for this patent, including its own cited references and the claim set.

Now let me examine the specific prior art references, including the JointPoint family cited in this space, and confirm the '219's own cited references.

Obviousness Analysis — U.S. Patent No. 11,883,219 B2 under 35 U.S.C. § 103

Builds on: the Patent Summary and Litigation Summary already generated above. I do not repeat the bibliographic table, the chain of title, or the litigation negative finding. Where the earlier sections flagged uncertainty, I carry those flags forward and add new ones.

0. Two threshold caveats before the merits

(a) I could not retrieve the "Prior Art" citation table. The full text of US 11,883,219 that was fetched on 2026‑09‑24 is truncated mid‑specification (it cuts off in the "INTRA‑OP REAMING FOR CUP PLACEMENT" passage). The Google Patents Citations / Cited By / Similar Documents tables therefore were not in the retrieved text. All I can confirm from the record itself is the "Prior art keywords" field: image, grid, pelvis, artificial intelligence, subject — which is itself an admission-shaped clue about where the examiner located the art (a grid + pelvis + AI combination).

Consequently, the combinations below are built from (i) art that surfaced as prior art in this search, and (ii) art the '219 specification itself cites or admits. This is not the same as a reproduction of the examiner's cited-art list. That distinction matters if you are preparing an IPR petition, because you must run the actual 1449/892 list.

(b) The claim text is reconstructed, not read. As noted in the Patent Summary, '219's abstract is word-for-word identical to claim 1 of same-family US 11,540,794 B2. I therefore analyze the claim scope as reconstructed (independent method claim ~1, independent system claim ~9, plus dependents 2–8 and 10). Any conclusion below is conditional on that reconstruction. If the granted '219 claims contain a limitation not present in the '794 claim set (e.g., a specific classifier-conditioning step), Combination A may not reach it.

1. The critical date — and a hard exclusion

Exclusion #1 — US 11,540,794 B2 (and US 2022/0265233 A1, US 2023/0000451 A1, WO 2020/056086 A1) are NOT available as prior art. These are same-family, same-inventive-entity, commonly-owned disclosures (Boddington, Saget, Cates, Oulhaj, Kubiak / OrthoGrid). They are not "by another" under §102(a)(2), and §103(b) / the common-ownership exception would bar their use in a §103 combination even if they were. Do not build a §103 ground on the '794 family. (This is a trap in this art unit family; the '219 and '794 abstracts are identical and are easy to mistake for separate art.)

Exclusion #2 — the recent clinical literature is NOT prior art, but is useful as background. In this search I retrieved, among others:

• the 2022 Hip Society Award paper (J. Arthroplasty) on U‑Net detection of teardrops, ischial wings/tuberosity, pubic symphysis, greater and lesser trochanters to compute HJC;
• Jang et al., J. Arthroplasty 38(10):2017‑2023 (2023) — DL-automated leg-length-discrepancy on AP pelvis radiographs using teardrop, obturator foramen, ischial tuberosity, greater/lesser trochanters;
• Mulford et al., Hip Int'l 36(3):388‑397 (2026) and Hip & Pelvis 38(3):234 (2026) — DL femoral/acetabular/global offset from AP pelvis radiographs, U‑Net, inter-teardrop line;
• Corin/Kavolus et al., AAOS 2024 — DL landmark detection on functional lateral radiographs (pubic symphysis, sacral endplate, ASIS) for spinopelvic measures.

Every one of these post-dates 2018‑09‑12 and is therefore unavailable as §102/§103 art. They are nonetheless powerful for the separate point that the claimed subject matter was the field's converging solution — but be disciplined and do not plead them as art. (Two of them are dated 2026; treat them as after‑the‑fact evidence only.)

2. The prior-art references that do qualify

Honesty flag: U‑Net (2015) and Payer et al. (2016) are references I know of but did not retrieve in this search session (I hit the tool-call budget). Their pre‑2018 dates are supplied from my own knowledge, not from a fetched source, so verify the exact citations before filing. The JointPoint, Penenberg, Matta, Baumgaertner, Bruijn and LeCun items were surfaced in this session.

Caveat on US 9,610,134: the title ("alignment plate … virtual alignment grid") is characteristic of OrthoGrid's own radiolucent hip-grid product, and the '219 specification incorporates it by reference. If US 9,610,134 is commonly owned with '219, it is excluded from a §103 combination by §103(b)/§102(b)(2)(C) on the same basis as the '794 family — but it remains usable as (i) evidence of the state of the art and (ii) an Applicant Admission by Incorporation of what a virtual alignment grid registered to an anatomic image was. Verify the real-party assignee before relying on it either way.

3. What each qualifying reference discloses, element by element

3.1 JointPoint (US 10,433,914 / US 2016/0100909 / US 2016/0128654) — the primary reference

JointPoint's abstract: "acquire (i) at least a reference image including one of a preoperative image of a surgical site … and a contralateral image on an opposite side of the patient from the surgical site, and (ii) at least an intraoperative image of the site after an implant has been affixed … generate at least one reference stationary point on at least the skeletal bone … and at least one intraoperative stationary point … The location of the implant is identified … including the position of first and second centers of rotation, which are digitally represented and copied into the reference image to analyze at least one of offset and length differential." (US 10,433,914 B2, abstract)

Specifically on point:

• Teardrop / pubic symphysis / obturator foramen as the landmark set. US 2020/0352529 A1 (JointPoint/Depuy continuation, whose specification traces to the 2015 filings) states: "different images of a pelvic girdle PG of a patient … typically show one or both obturator foramen OF and a central pubic symphysis PS, which the present inventors have recognized as suitable reference points or features for use as part of a stationary base … Other useful anatomical features … include femoral neck FN and lesser trochanter LT … femoral head FH and greater trochanter GT … ischial tuberosities IT … a 'tear drop' TD relating to a bony ridge along the floor of the acetabular fossa, and the anterior superior iliac spine ASIS and the anterior inferior iliac spine AIIS." (US 2020/0352529 A1 ¶¶0049–0051)
• Grid-based comparison of the good and operative sides. "reference (preop) and intraop images are compared via a grid-type X‑Y coordinate system." And: "a stationary base line begins at the pubic symphysis PS, touches or intersects at least a portion of an obturator foramen OF, and extends to (i) the 'tear drop' TD, or (ii) the anterior inferior iliac spine AIIS."
• Confidence scoring of the overlay/match. "the present Image Overlay technique can analyze how 'similar' these images are to give the user feedback as to how accurate the results are, that is, to provide a confidence interval"; "the system provides a weighted 'confidence score', ranging from 0.0 to 10.0 … a cut-off value can be provided beyond which the error is deemed to be too great … the system then recommends that the user obtain an alternative intraoperative image to compare with the preoperative image, or with a contralateral image." (US 2016/0128654 A1)
• Vector decomposition into leg-length and offset — squarely the dependent-claim subject matter. "the Calculation Module … calculates the change in leg length and offset by analysing the vector between the greater trochanter point in the preop image and the calculated phantom point in the intraop image … To calculate leg length, the system calculates the distance between these two points along the femoral axis … To calculate offset, the system calculates the distance between the two points along the axis that is perpendicular to the femoral axis." (same source)
• Implant/hardware identification and scaling. "identifying the prosthetic implant device itself while entering the known metric size for at least one dimension of the device"; "Draw Circle Around Acetabular Component and Enter Known Size of Component."
• The manual-bottleneck admission that supplies the motivation. JointPoint's own workflow steps are user-driven: "the user is guided to identify one or more landmark points (i.e. the tear drop anatomical feature of the pelvis) on each image." (US 2015/0238271 A1)

3.2 US 9,610,134 B2 — the virtual alignment grid

The '219 specification incorporates it expressly in the context of the best-fit transformation: "The best fit transformation is computed using a variety of established methods … See, e.g., U.S. Pat. No. 9,610,134 specifically incorporated by reference in its entirety." Its title — "Alignment plate apparatus and system and method of use with virtual alignment grid" — discloses a physical alignment plate bearing a radiopaque grid, and the corresponding virtual grid overlaid on the anatomic image. (US 9,610,134 B2)

3.3 Penenberg and the clinical literature — measurement definitions

• Penenberg (US 2014/0378828 A1; US 8,831,324 B2) — intra‑op hip analysis apparatus considered in JointPoint's Background alongside the Matta technique.
• Matta et al. 2005 — the manual AP-pelvis overlay technique for cup position and leg length: the exact manual workflow the '219 automates.
• Baumgaertner 1995 / Bruijn 2012 — tip-apex distance and screw cut-out predictors: the "known predictors of complications" that the '219's "Failure Risk Score" module formalizes.

3.4 Deep-learning art for image landmarking

• LeCun, Bengio & Hinton, Nature 2015, is cited on the face of the '219 specification (in the AI Engine passage): "Any number of classical machine learning approaches can be used, as well as more modern Deep learning networks [LeCun, Yann, Yoshua Bengio, and Geoffrey Hinton. 'Deep learning.' nature 521.7553 (2015): 436] … such as Convolutional Neural Networks." This is an Applicant Admission that CNN-based image classification was known and available.
• U‑Net (2015) / Payer et al. (2016) — encoder–decoder CNNs and CNN heat-map regression specifically for anatomical landmark localization on medical images (see honesty flag in §2).

4. The § 103 combinations

Combination A (the primary ground) — JointPoint + CNN landmark localization

Combine: US 10,433,914 B2 (and its published predecessors US 2016/0100909 A1, US 2016/0128654 A1, US 2015/0238271 A1) + LeCun 2015 / Ronneberger U‑Net 2015 / Payer 2016 (CNN landmark regression).

Result: renders reconstructed independent claim 1 obvious.

The only gap is the neural network. Everything else — good-side-referenced landmark detection, the teardrop/symphysis landmark set, the grid, leg length, offset, hardware — is in JointPoint.

Combination B — JointPoint + US 9,610,134 (virtual alignment grid) + CNN

This is Combination A with US 9,610,134 supplying the express "virtual alignment grid" registered to an anatomic image, which forecloses any argument that the "functional pelvis grid" was novel as a grid. Note the common-ownership risk in §2 before pleading it.

Combination C (system claim 9) — Combination A/B + conventional layered software architecture

Claim 9 recites a data layer / algorithm layer / application layer software module. Layered (presentation–logic–data) software architecture was, by 2018, the ordinary design pattern; the '219 specification's own text refers to it with no asserted unexpected result. Reciting a known technique (layering) applied to a known system (JointPoint's overlay) with no new and unexpected result is the paradigm of KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007). Expect this limitation to be treated as an insignificant design choice.

Combination D (dependents 5–7) — essentially anticipated by JointPoint

Reconstructed dependent claims 5 (quantify measurements; calculate a user-requested output from vector or vertical distances), 6 (leg length), and 7 (offset) read on JointPoint's express disclosure: leg length = distance along the femoral axis; offset = distance along the perpendicular axis between the greater-trochanter point and the phantom point. That is a literal vector-decomposition disclosure. If the '219 dependent claims are as reconstructed, they are anticipated under §102(b)(1), and a fortiori obvious — no combination needed.

Combination E (dependents 2–4, 8, 10)

• Cl. 2 (register good side → operative side): JointPoint's overlay of contralateral and operative images using the pelvic reference line.
• Cl. 3 (register pre‑op ipsilateral → post‑op ipsilateral): JointPoint's "Reverse Templating" and preop/postop overlay flowcharts.
• Cl. 4 / cl. 10 (anatomy map of an entire anatomical region): JointPoint's multi-landmark stationary base spanning PS, OF, TD, ASIS, AIIS, ischial tuberosity, GT, LT — a region-wide landmark map.
• Cl. 8 (detect hardware by classifying against the good side): JointPoint's acetabular-component identification and known-size scaling, plus the '219's own admission that hardware detection is a recognition task.

5. Motivation to combine (the KSR / In re Rouffet analysis)

A POSITA at 2018‑09‑12 — a medical-image computing engineer or a computer-assisted-orthopedics engineer with hip-arthroplasty domain exposure — would have had at least six independent, articulated motivations:

1. JointPoint's own express admission of the manual bottleneck. JointPoint's entire workflow is user-guided landmark identification: "the user is guided to identify … the tear drop anatomical feature of the pelvis on each image." Automating a repeatedly-performed human identification step with the CNN tools that LeCun 2015 (cited by Applicant) and the 2015–2016 U‑Net/Payer literature made routine is the most natural possible improvement. Motivation: speed, reproducibility, inter-rater variability, and OR throughput.
2. Same field, same problem, same images. JointPoint and the CNN-landmark art both operate on AP pelvis radiographs/fluoroscopy; JointPoint addresses operative-vs-contralateral hip comparison; the CNN art addresses landmark localization on radiographs. KSR "familiar elements according to known methods."
3. Predictable result. Detecting a teardrop or the pubic symphysis is a bounded, well-defined landmark-localization task with a ground-truth definition already established in the clinical literature (Baumgaertner 1995; Matta 2005) — a POSITA would expect the CNN to succeed (and, as the post-2018 literature confirms, it did). KSR: "the improvement is a predictable use of prior-art elements according to their established functions."
4. A finite, identified, and small landmark set. The art had already enumerated the candidate landmarks (TD, PS, OF, LT, GT, IT, ASIS, AIIS). Nothing remained but to select and detect them.
5. Design incentive to remove intra-op steps. JointPoint's stated purpose is to reduce outliers without navigation hardware; automating landmarking advances that same purpose rather than working against it.
6. Commercial/clinical pressure. The '219 Background itself frames the problem: "the position of the patient's pelvis varies widely during surgery and from patient to patient" and "the placement of an implant or the reduction of a bone fragment can still not be as close as desired by the surgeon." That is an articulated problem statement in the patent, which cuts against non-obviousness.

No teaching away. JointPoint's criticism of CT-based computer-assisted navigation (cost, learning curve, hardware insertion) is a criticism of navigation, not of automated image analysis. If anything, JointPoint's stated goal — navigation-grade accuracy without navigation hardware — directs the artisan toward software-only automation of the landmark step.

No unexpected results pleaded. The '219 specification asserts improved alignment generically but does not tie any measured, comparative, unexpected result to the specific combination of good-side-referenced CNN landmark detection + functional pelvis grid.

6. Rebuttals the patent owner will raise, and how they fare

7. Conclusion

• Reconstructed independent claim 1 is obvious over Combination A: US 10,433,914 B2 / US 2016/0128654 A1 / US 2016/0100909 A1 (JointPoint) in view of LeCun et al., Nature 521:436 (2015) — which the '219 specification itself cites — and the 2015–2016 CNN medical-landmark literature. JointPoint supplies every limitation except a neural-network implementation of detection; the CNN art supplies that; the motivation is JointPoint's own manual-landmarking bottleneck and the field's stated reproducibility problem.
• Reconstructed independent claim 9 is obvious over Combination A plus the conventional three-layer software architecture (Combination C).
• Reconstructed dependents 5, 6, and 7 (vector/vertical distances; leg length; offset) are, on this record, anticipated by JointPoint's express vector-decomposition teaching, and obvious in any event.
• Reconstructed dependents 2–4, 8, and 10 are obvious over JointPoint alone or in view of US 9,610,134.
• Confidence: moderate. Two independent limits apply: (i) the claim text is reconstructed from same-family US 11,540,794, not read from the '219 grant; and (ii) I was unable to retrieve the actual "Prior Art"/Citations table from the truncated Google Patents fetch, so I cannot certify that the examiner considered JointPoint. If JointPoint does not appear on the '219 face, its absence is itself relevant to the strength of an IPR ground.

Two follow-ups worth running on the file

1. Pull the real '219 claim set (and the '219 file wrapper / 892 list) — the whole analysis pivots on it.
2. Check the lead flagged in the Litigation Summary: USPTO proceeding IPR2025‑01027 and the consolidated E.D. Tex. action (consolidated 2025‑02‑18, trial 2026‑09‑14). If either targets this family, the petitioner's ground of unpatentability will name the actual art and claim construction — which would supersede my reconstruction. I could not confirm the patent number in that proceeding.

Nothing above should be read as a legal conclusion or as a certified invalidity opinion; it is a technical prior-art/obviousness assessment subject to the two confidence limits stated.