US 10991097

US Patent 10991097, titled "Artificial intelligence segmentation of tissue images," was issued to Tempus AI Inc. The patent describes methods and systems for analyzing digital medical images, particularly histological slides of cancerous tissue, using artificial intelligence.

Here is a summary of the patent details:

• Title: Artificial intelligence segmentation of tissue images
• Assignee: Tempus AI Inc (current assignee as of February 9, 2024, via change of name from Tempus Labs Inc)
• Inventors: Stephen Yip, Irvin Ho, Lingdao Sha, Boleslaw Osinski
• Filing Date: December 31, 2019
• Issue Date: April 27, 2021
• Abstract: "Techniques for generating an overlay map on a digital medical image of a slide are provided, and include cell detection and tissue classification processes. Techniques include receiving a medical image, separating the image into tiles, and performing tile classifications and tissue classifications based on a multi-tile analysis. Techniques additionally include identifying cell objects in the image, separating the image into and displaying polygons identifying the cell objects and cell classifications. Generated displays may be overlays over the initial digital image."

Independent Claims Overview:

Independent Claim 1:
This claim describes a computer-implemented method for creating an overlay map on a digital image of a slide. The method involves receiving a digital image, dividing it into multiple tiles, and then identifying the most prevalent tissue class within each tile using an analysis that considers multiple tiles.

Independent Claim 11:
This claim describes a computer-implemented method for classifying tissue in a digital image of a slide. It involves receiving the digital image, generating a digital overlay drawing specifically for a tissue region within that image, and then displaying tiles over this tissue region. The method visually identifies the predicted content of each tile to create a classification map of the digital image.

Independent Claim 14:
This claim describes a computer-implemented method for detecting cells in a digital image of a slide. The method receives the digital image, identifies specific cell objects within it, generates a digital overlay drawing for the image, and then displays a polygon outline around each of the identified cell objects in this overlay.

Independent Claim 17:
This claim describes a computer-implemented method for classifying tissue in a digital image of a slide that combines both tile and cell-level analysis. It involves receiving the digital image, segmenting it into tiles, and predicting a class for each tile. Simultaneously, it identifies multiple cell objects in the image and predicts a class for each cell object. Crucially, for any tile that corresponds to an identified cell object, the method assigns the predicted class of that specific cell object to the tile, overriding the tile's initial predicted class.

Independent Claim 27:
This claim describes a system that performs the methods outlined in independent claim 1. It comprises a processor and a computer-readable memory storing instructions. When executed by the processor, these instructions cause the system to receive a digital image, divide it into tiles, and identify the majority tissue class within each tile based on a multi-tile analysis.

Independent Claim 30:
This claim describes a system that performs the methods outlined in independent claim 11. It includes a processor and a computer-readable memory storing instructions. When executed, these instructions enable the system to receive a digital image, generate an overlay drawing of a tissue region, and display tiles over that region, visually identifying the predicted content of each tile to generate a classification map.

Independent Claim 31:
This claim describes a system that performs the methods outlined in independent claim 14. It comprises a processor and a computer-readable memory storing instructions. When executed, these instructions cause the system to receive a digital image, identify cell objects, generate a digital overlay drawing, and display polygons outlining each identified cell object.

Independent Claim 32:
This claim describes a system that performs the methods outlined in independent claim 17. It includes a processor and a computer-readable memory storing instructions. When executed, these instructions enable the system to receive a digital image, segment it into tiles, determine a predicted class for each tile, identify cell objects, determine a predicted class for each cell object, and for corresponding tiles and cell objects, assign the cell object's predicted class to the tile.

Legal Status and Litigation:

As of April 26, 2026, the patent US10991097B2 is listed as "Active" with an anticipated expiration date of December 31, 2039.

There are several litigation activities noted:

• A PTAB case, IPR2026-00185, was filed and is currently pending.
• The first worldwide family litigation was filed.
• US cases were filed in the California Northern District Court (cases 3:25-cv-06622 and 5:25-cv-06622).
• A US case was filed in the California Southern District Court (case 3:25-cv-00621).

There is no mention of specific CAFC 2026 dockets in the provided patent information. The pending IPR case would typically be heard by the PTAB (Patent Trial and Appeal Board), and the district court cases are at the district level. Appeals from these bodies would go to the CAFC, but no specific CAFC dockets for 2026 are explicitly listed as active for this patent in the provided text.

As of April 26, 2026, the following litigation involving US Patent 10,991,097 is known:

• Plaintiff(s): Tempus AI Inc
• Defendant(s): Guardant Health Inc
• Jurisdiction: U.S. District Court for the Southern District of California
• Case Number: 3:25-cv-00621
• Filing Date: March 14, 2025
• Outcome/Current Status: The case is for patent infringement, and a first-instance judgment was issued on August 1, 2025, with an outcome of "other."

Additionally, Unified Patents has filed an inter partes review (IPR) case against US Patent 10,991,097:

• Plaintiff(s) (Petitioner): Unified Patents
• Defendant(s) (Patent Owner): Tempus AI Inc (current assignee of US10991097)
• Jurisdiction: Patent Trial and Appeal Board (PTAB)
• Case Number: IPR2026-00185
• Filing Date: Not explicitly stated, but the status is "Pending" with a critical flag.
• Outcome/Current Status: Pending.

Proceedings overview

There is one AIA trial proceeding on file for US Patent 10991097, IPR2026-00185. This proceeding resulted in a discretionary denial of institution, meaning the PTAB chose not to initiate a full review of the patent's claims. For a defendant, this indicates that the patent has not been subjected to a full IPR trial, and its claims remain intact. An IPR-based defense on the specific grounds raised in this petition would be more challenging due to the discretionary denial precedent.

IPR2026-00185 — Guardant Health, Inc. v. Tempus AI Inc.

• Type: Inter Partes Review
• Filed: 2025-12-30
• Status: Discretionary Denial (The PTAB declined to institute the inter partes review, meaning the trial did not proceed to a merits determination)
• Judge panel: Not publicly available from the initial search, as the case was denied institution before a full panel would typically be assigned for a trial.
• Petition grounds: Details not publicly available from the initial search results, but a discretionary denial implies the grounds were evaluated for institution.
• Institution decision: Denied on 2026-04-28. The denial was discretionary. This often occurs when the PTAB applies its precedential guidance, such as the Fintiv factors, to decline institution even if the merits threshold (reasonable likelihood of unpatentability) might be met. The reasoning for a discretionary denial typically involves factors like parallel district court litigation, trial schedule, or stage of the litigation.
• Final Written Decision (if issued): Not issued, as institution was denied.
• Settlement / termination: The proceeding was terminated via discretionary denial of institution. There was no settlement recorded in the provided information.
• Appeal: No appeal to the Federal Circuit, as there was no Final Written Decision on the merits.
• Defensive value: The discretionary denial of institution in IPR2026-00185 means that none of the claims of US10991097 were challenged on the merits and found unpatentable by the PTAB. For a defendant, this implies that the specific prior art and arguments raised by Guardant Health, Inc. in this petition were not deemed sufficient by the PTAB to proceed to trial, likely due to procedural or efficiency considerations rather than a full review of patentability. Future IPRs against this patent would need to present stronger or different arguments, or address the factors that led to the discretionary denial.

Strategic summary

Currently, all claims of US10991097 remain SUSTAINED as no AIA trial has proceeded to a Final Written Decision invalidating any claims. IPR2026-00185 was denied institution on discretionary grounds, meaning the PTAB did not assess the patentability of the claims on their merits. Consequently, the patent has not been narrowed through IPR proceedings.

Regarding the estoppel landscape, since IPR2026-00185 was denied institution, § 315(e)(2) estoppel, which bars petitioners (and their privies) from raising any ground they raised or reasonably could have raised in a final written decision, would not apply here. This is because there was no Final Written Decision. Therefore, the prior-art grounds presented in that petition could theoretically still be raised by Guardant Health or its privies in district court or other proceedings, though the PTAB's discretionary denial might suggest weaknesses in the petition's overall strategy or timing in relation to parallel litigation. For a defendant currently being asserted against, all prior-art grounds remain available, as no claims have been formally tested and affirmed as patentable in a full IPR trial.

The pattern signals indicate that Guardant Health, Inc. has initiated an IPR against this patent. The discretionary denial suggests that the PTAB is actively applying its discretion, potentially to manage parallel litigation or for other policy reasons. This signals that subsequent IPRs might face similar discretionary hurdles if not carefully strategized to differentiate from the circumstances of IPR2026-00185.

Recommended next steps

• Review the institution denial decision for IPR2026-00185 (Unified Patents has a link on their portal) to understand the specific reasons for the discretionary denial. This will be crucial for any future IPR strategy.
• Given that all claims are currently sustained, any defendant facing assertion should conduct a thorough prior art search to develop robust invalidity contentions, as the patent has not been "hardened" by surviving a full IPR trial on the merits.
• Monitor the district court litigation cases (3:25-cv-06622 and 5:25-cv-06622 in California Northern District Court, and 3:25-cv-00621 in California Southern District Court) for any developments that might shed light on the claims or potential settlement strategies.
• If considering filing a new IPR, carefully assess the PTAB's discretionary denial factors (e.g., Fintiv factors) and tailor the petition to minimize the risk of a similar denial.

Inventors

• Stephen Yip (Tempus Labs Inc)
• Irvin Ho (Tempus Labs Inc)
• Lingdao Sha (Tempus Labs Inc)
• Boleslaw Osinski (Tempus Labs Inc)

All inventors were employees of the original assignee, Tempus Labs Inc, at the time of filing.

Original assignee

The original assignee, Tempus Labs Inc, is an operating company focused on artificial intelligence and precision medicine, including molecular and clinical data analysis for oncology. They ship products embodying the claims, specifically AI-powered diagnostic and treatment selection tools that analyze tissue images. Tempus Labs Inc changed its name to Tempus AI, Inc. on February 9, 2024, and continues to operate under that new name.

Assignment timeline

• 2020-08-03 (executed) / recorded 2020-08-03 — Reel 053913/0890
  • Conveyance: Assignment
  • Assignor: YIP, STEPHEN; OSINSKI, Boleslaw; HO, Irvin; SHA, Lingdao
  • Assignee: TEMPUS LABS, INC.
  • Correspondent: MCDONNELL BOEHNEN HULBERT & BERGHOFF LLP, 300 SOUTH WACKER DRIVE, SUITE 3200, CHICAGO, ILLINOIS, UNITED STATES, 60606
  • Context: Transfer of inventor interests to the original assignee.
• 2022-09-22 (executed) / recorded 2022-09-22 — Reel 061144/0137
  • Conveyance: Security Interest
  • Assignor: TEMPUS LABS, INC.
  • Assignee: ARES CAPITAL CORPORATION, AS COLLATERAL AGENT
  • Correspondent: SIMPSON THACHER & BARTLETT LLP, 425 LEXINGTON AVENUE, NEW YORK, NEW YORK, UNITED STATES, 10017
  • Context: Securitization of assets, likely for financing.
• 2024-02-09 (executed) / recorded 2024-02-09 — Reel 066744/0785
  • Conveyance: Change of Name
  • Assignor: TEMPUS LABS, INC.
  • Assignee: TEMPUS AI, INC.
  • Correspondent: MCDONNELL BOEHNEN HULBERT & BERGHOFF LLP, 300 SOUTH WACKER DRIVE, SUITE 3200, CHICAGO, ILLINOIS, UNITED STATES, 60606. This correspondent also appears on the 2020-08-03 assignment.
  • Context: Corporate name change.

Timeline diagram

timeline
    title Ownership of US 10991097
    2019 : Filed by Tempus Labs Inc
    2020 : Inventor Assignment to Tempus Labs Inc
    2021 : Issued to Tempus Labs Inc
    2022 : Security Interest to Ares Capital Corp
    2024 : Name Change to Tempus AI Inc
    2025 : First infringement suit filed
    2026 : IPR filed by Unified Patents

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The assignees (Tempus Labs Inc / Tempus AI, Inc. and Ares Capital Corporation) are recognized operating companies or financial institutions.
2. Known asserter in the chain — not present. None of the assignees (Tempus Labs Inc / Tempus AI, Inc., Ares Capital Corporation) are known NPEs. Unified Patents is a petitioner in an IPR, which is an anti-NPE action, not an asserter.
3. Repeat correspondent across the chain — present. MCDONNELL BOEHNEN HULBERT & BERGHOFF LLP appears as the correspondent on both the 2020-08-03 assignment (Reel 053913/0890) and the 2024-02-09 change of name (Reel 066744/0785).
4. Cascading transfers — not present. There are no multiple consecutive assignments through chained LLCs within a short timeframe.
5. Pre-litigation transfer — not present. The assignment to Tempus Labs Inc from the inventors occurred in August 2020. The first infringement suit was filed in March 2025. The security interest filing in September 2022 and the name change in February 2024 are not indicative of pre-litigation transfer for assertion.
6. Bankruptcy fire-sale — not present. There is no indication of bankruptcy proceedings for Tempus Labs Inc or Tempus AI, Inc.
7. Privateering — unclear. There is no information in the provided patent text or assignment records to suggest a privateering arrangement.
8. Defensive aggregator (anti-NPE) — present. Unified Patents has filed an IPR against the patent, indicating an action taken by a defensive aggregator to challenge the patent's validity.

Verdict

Operating-company assertion. The patent is currently held by Tempus AI, Inc., which is an operating company that develops products related to the patent's claims. While Unified Patents (a defensive aggregator) has filed an IPR, this action challenges the patent rather than indicating an NPE assertion. The repeated correspondent for Tempus-related filings suggests internal consistency rather than a shell-entity pattern. The litigation filed in district courts by Tempus AI, Inc. (as Plaintiff) against Guardant Health Inc further supports this, as Guardant Health Inc is a competitor in the precision medicine and diagnostics space.

USPTO Assignment Center search page: https://assignmentcenter.uspto.gov/

To identify the most relevant prior art for US Patent 10991097, I will search the USPTO database for the patent and then examine its cited references. The USPTO's Patent Public Search tool is suitable for this purpose, offering access to full text for patents and prosecution history which includes prior art citations.

It is important to note that a comprehensive prior art analysis can be quite extensive, involving not only cited patents but also non-patent literature, foreign patents, and other public disclosures. For this task, I will focus on the patent citations listed within the 10991097 patent document itself.

I will now proceed to search the USPTO database for patent 10991097.

To identify the most relevant prior art for US Patent 10991097, I accessed the patent document on Google Patents, which directly provides the "Prior art citations" section. This section lists the patents and other publications that the examiner and applicant considered during the prosecution of US10991097.

Below is a breakdown of the most relevant prior art cited in US Patent 10991097, along with their publication/filing dates, brief descriptions, and potential anticipation of claims under 35 U.S.C. § 102. It is important to note that a full anticipation analysis would require a detailed claim-by-claim comparison, but I can provide a high-level assessment based on the descriptions.

Patent Citations

• US9619864B2 (Methods and systems for cell detection and analysis in digital images)
  • Publication Date: April 11, 2017
  • Description: This patent describes methods and systems for cell detection and analysis in digital images, which involve identifying cell objects and extracting features for analysis.
  • Potential Anticipated Claims: This patent potentially anticipates aspects of Independent Claim 14 (method for cell detection, identifying cell objects, generating and displaying polygons outlining cells) and Independent Claim 31 (system for cell detection).
• US9477969B2 (Method and system for image processing)
  • Publication Date: October 25, 2016
  • Description: This patent generally relates to image processing methods and systems. Without further detail, it's difficult to pinpoint specific anticipation. However, given the context of US10991097, it likely covers general image processing techniques that could be applied to medical images.
  • Potential Anticipated Claims: Depending on the specifics of its image processing techniques, it could broadly relate to elements of Independent Claim 1 (receiving digital image, separating into tiles), Independent Claim 11 (generating overlay drawing, displaying tiles), Independent Claim 17 (segmenting digital image into tiles), and their corresponding system claims (27, 30, 32).
• US10235777B2 (Systems and methods for automated analysis of microscopy images using deep learning)
  • Publication Date: March 19, 2019
  • Description: This patent focuses on automated analysis of microscopy images, specifically using deep learning techniques. This is highly relevant to the core of US10991097.
  • Potential Anticipated Claims: This patent strongly anticipates Independent Claim 1 (multi-tile analysis for majority tissue class), Independent Claim 11 (tissue classification using a classification model, potentially multi-resolution FCN), Independent Claim 17 (segmenting into tiles, determining predicted class for each tile), and their corresponding system claims (27, 30, 32), particularly concerning the use of deep learning for image analysis and classification in medical microscopy.
• US20180247348A1 (Analyzing histology images with deep learning)
  • Publication Date: August 30, 2018
  • Description: This application describes analyzing histology images with deep learning. This is directly in line with the subject matter of US10991097.
  • Potential Anticipated Claims: Similar to US10235777B2, this application strongly anticipates aspects of Independent Claim 1, 11, and 17, and their corresponding system claims (27, 30, 32), especially regarding deep learning for histology image analysis and tissue classification.

Non-Patent Literature

No specific non-patent literature references were provided in the initial patent text, so I am unable to analyze them. A full prior art search would typically include a review of scientific papers, articles, and other publications.

The obviousness of US patent 10991097 under 35 U.S.C. § 103 can be assessed by considering combinations of the prior art explicitly mentioned or acknowledged within the patent document itself. A person having ordinary skill in the art (POSA) in the field of AI-driven medical image analysis would be motivated to combine these known techniques to address existing challenges in histological image segmentation.

Common General Knowledge and Explicit Prior Art:

The patent US10991097 acknowledges several existing technologies and challenges in the background and detailed description:

1. Convolutional Neural Networks (CNNs): Acknowledged as deep learning algorithms that analyze digital images by assigning one class label to each input image. The patent notes their limitation for slides containing multiple tissue types [cite: "A Convolutional Neural Network (“CNN”) is a deep learning algorithm that analyzes digital images by assigning one class label to each input image. Slides, however, include more than one type of tissue, including the borders between neighboring tissue classes."].
2. Fully Convolutional Networks (FCNs): Acknowledged as capable of assigning classification labels to each pixel within an image, thus more useful for images with multiple classifications and generating overlay maps. However, the patent states that traditional FCNs for digital slides are impractical due to the extensive annotation time and computational requirements for pixel-level labeling of high-resolution images [cite: "A Fully Convolutional Network (FCN) can analyze an image and assign classification labels to each pixel within the image, so a FCN is more useful for analyzing images that depict objects with more than one classification.", "However, FCN deep learning algorithms that analyze digital slides would require training data sets of images with each pixel labeled as a tissue class, which requires too much annotation time and processing time to be practical."].
3. ResNet-18 Image Recognition Model: Explicitly mentioned as a "known CNN" that forms the basis for the tile-resolution FCN PhiNet architecture described in the patent [cite: "The tissue class locator 216 includes a tile-resolution fully convolutional network (FCN) black box deep learning model based on a known CNN ResNet-18 image recognition model.", "FIG. 6B illustrates the differences between the ResNet-18 algorithm on the left, and the tile-resolution FCN PhiNet shown in FIG. 6A and on the right half of FIG. 6B."].
4. Two-class UNet Models with Binary Classification: Acknowledged as "known in the art" for cell segmentation (e.g., cell vs. background). The patent highlights their limitation in accurately counting overlapping cells [cite: "Two-class UNet models with binary classification (cell vs background) are known in the art, but the three-class UNet model allows a type of classification that is not binary, which requires adaptation with the use of a different loss function.", "In traditional two-class cell outlining models that only label whether a pixel contains a cell outer edge or not, each clump of two or more overlapping cells would be counted as one cell."].
5. U.S. Provisional Patent Application No. 62/889,521: This provisional application, titled "Determining Therapeutic Tumor-Infiltrating Lymphocytes (TILS) from Histopathology Slide Images," filed on August 20, 2019, is incorporated by reference, indicating its status as prior art for the present patent (filed December 31, 2019) [cite: "An example of a TILS process and engine is disclosed, for example, in U.S. Provisional Patent Application No. 62/889,521, titled “Determining Therapeutic Tumor-Infiltrating Lymphocytes (TILS) from Histopathology Slide Images,” filed on Aug. 20, 2019, which is incorporated herein by reference"].

Obviousness Combinations and Motivation:

The independent claims of US10991097 generally cover methods and systems for:

• Tile-based tissue classification using multi-tile analysis (Claims 1, 11, 27, 30).
• Cell detection and outlining using polygons (Claims 14, 31).
• Combining tile and cell classifications, with cell classification overriding tile classification (Claims 17, 32).

A POSA would have been motivated to combine the known prior art to achieve the claimed inventions for the following reasons:

Combination 1: ResNet-18 + Traditional FCNs → Tile-resolution FCN (PhiNet) for Tissue Classification (Claims 1, 11, 27, 30)

• Rationale: The patent explicitly states the limitations of traditional CNNs (one label per image) and traditional FCNs (computational intensity for pixel-level segmentation of large images) for analyzing complex histological slides with diverse tissue types [cite: "There is a need to classify different regions as different tissue classes, in part to study the borders between neighboring tissue classes and the presence of immune cells among tumor cells.", "The high number of pixels makes it infeasible to use traditional FCNs to segment digital images of slides."].
• Motivation: A POSA, faced with these known problems, would be motivated to adapt existing, proven deep learning architectures like ResNet-18 to enable efficient, multi-class segmentation of large digital histology images. The adaptation to a "tile-resolution" FCN (PhiNet) with "additional layers" for a "classification-segmentation task" on tiles, as described in the patent, directly addresses the computational and annotation challenges of pixel-level FCNs while providing more granular classification than a whole-image CNN [cite: "the added layers convert a classification task into a classification-segmentation task. This means that instead of receiving and classifying a whole image as one tissue class label, the added layers allow the tile-resolution FCN to classify each small tile in the user-defined grid as a tissue class."]. The concept of dividing an image into tiles for processing is a standard image processing technique, and applying an adapted convolutional network for tile-level classification would be an obvious step for a POSA seeking to balance computational efficiency with localized analysis. The "multi-tile analysis" and context-awareness (medium tiles providing context for small central tiles) described in the patent would be an expected refinement to improve accuracy by leveraging local neighborhood information, a common approach in image recognition.

Combination 2: Two-class UNet Models + General AI Image Segmentation Principles → Three-class UNet Model for Cell Detection (Claims 14, 31)

• Rationale: The patent acknowledges that "Two-class UNet models with binary classification (cell vs background) are known in the art" [cite: "Two-class UNet models with binary classification (cell vs background) are known in the art, but the three-class UNet model allows a type of classification that is not binary, which requires adaptation with the use of a different loss function."]. The core problem identified is the inaccuracy in counting individual cells when they overlap, a critical issue for quantitative analysis like Tumor-Infiltrating Lymphocyte (TIL) assessment [cite: "This facilitates the counting of each individual cell, especially when two or more cells overlap each other. In one example, tumor infiltrating lymphocytes will overlap tumor cells. In traditional two-class cell outlining models that only label whether a pixel contains a cell outer edge or not, each clump of two or more overlapping cells would be counted as one cell."].
• Motivation: A POSA specializing in biomedical image analysis would be motivated to improve the precision of cell detection and counting, especially for overlapping cells. Modifying a known two-class UNet to a three-class model (background, cell outer edge, cell interior) to explicitly distinguish cell boundaries from interiors would be an obvious solution to this known problem of overlapping cells. This adaptation would be driven by the clear need for more accurate individual cell segmentation, and the patent itself details this adaptation, including the need for a "different loss function" [cite: "Two-class UNet models with binary classification (cell vs background) are known in the art, but the three-class UNet model allows a type of classification that is not binary, which requires adaptation with the use of a different loss function."].

Combination 3: Tile-resolution FCN (PhiNet) for Tissue Classification + Three-class UNet for Cell Detection + U.S. Provisional Patent Application No. 62/889,521 → Integrated System with Cell-overriding-Tile Classification (Claims 17, 32)

• Rationale: Independent claims 17 and 32 describe an integrated system where both tile-level tissue classes and cell objects are identified, and the cell object's predicted class overrides the tile's predicted class if they correspond. This represents a hierarchical approach to segmentation.
• Motivation: A POSA aiming to provide a comprehensive and highly accurate diagnostic tool would be strongly motivated to integrate different levels of AI analysis. The limitations of tile-level classification (a tile might contain multiple cell types, making a single majority class assignation potentially misleading) would lead a POSA to seek refinement. The detailed, pixel-level information from a cell detection model (like the three-class UNet) offers this refinement. Combining the tile-based tissue classification (from PhiNet, as in Combination 1) with the individual cell detection and classification (from the UNet, as in Combination 2) and allowing the more precise cell-level data to override broader tile-level classifications is a logical and obvious step to enhance diagnostic accuracy. This is especially true given the clinical importance of specific cell types, such as Tumor-Infiltrating Lymphocytes (TILs), which are explicitly mentioned in the patent as being detectable by the system and taught in the incorporated provisional application [cite: "the digital tissue segmenter 201 is configured to determine tumor-infiltrating lymphocytes (TILS). An example of a TILS process and engine is disclosed, for example, in U.S. Provisional Patent Application No. 62/889,521, titled “Determining Therapeutic Tumor-Infiltrating Lymphocytes (TILS) from Histopathology Slide Images,” filed on Aug. 20, 2019, which is incorporated herein by reference"]. A POSA would understand that a specific cell type identified by a highly accurate cell model should take precedence over a coarser tissue classification for a region, particularly in contexts like predicting immunotherapy response. The integration would create a more robust and clinically relevant system.

In summary, the advancements described in US10991097, while useful, appear to be a logical progression for a person having ordinary skill in the art, driven by well-understood problems in medical image analysis and building upon existing, known AI architectures and techniques. The motivations for combining these elements—namely, to improve efficiency, accuracy, and clinical relevance in the analysis of complex histological images—would have been readily apparent.