Evaluating Vision and Pathology Foundation Models for Computational Pathology: A Comprehensive Benchmark Study.

To advance precision medicine in pathology, robust AI-driven foundation models are increasingly needed to uncover complex patterns in large-scale pathology datasets, enabling more accurate disease detection, classification, and prognostic insights. However, despite substantial progress in deep learning and computer vision, the comparative performance and generalizability of these pathology foundation models across diverse histopathological datasets and tasks remain largely unexamined. In this study, we conduct a comprehensive benchmarking of 31 AI foundation models for computational pathology, including general vision models (VM), general vision-language models (VLM), pathology-specific vision models (Path-VM), and pathology-specific vision-language models (Path-VLM), evaluated over 41 tasks sourced from TCGA, CPTAC, external benchmarking datasets, and out-of-domain datasets.

Evaluating Vision and Pathology Foundation Models for Computational Pathology: A Comprehensive Benchmark Study.

To advance precision medicine in pathology, robust AI-driven foundation models are increasingly needed to uncover complex patterns in large-scale pathology datasets, enabling more accurate disease detection, classification, and prognostic insights. However, despite substantial progress in deep learning and computer vision, the comparative performance and generalizability of these pathology foundation models across diverse histopathological datasets and tasks remain largely unexamined. In this study, we conduct a comprehensive benchmarking of 31 AI foundation models for computational pathology, including general vision models (VM), general vision-language models (VLM), pathology-specific vision models (Path-VM), and pathology-specific vision-language models (Path-VLM), evaluated over 41 tasks sourced from TCGA, CPTAC, external benchmarking datasets, and out-of-domain datasets.

Distinct gene signatures define the epithelial cell features of mucinous appendiceal neoplasms and pseudomyxoma metastases.

Introduction: Appendiceal mucinous neoplasms (AMN) are rare tumors of the gastrointestinal tract. They metastasize with widespread abdominal dissemination leading to pseudomyxoma peritonei (PMP), a disease with poor prognosis. There are many unknowns about the cellular features of origin, differentiation and progression of AMN and PMP.

A spatial transcriptomic signature of 26 genes resolved at single-cell resolution characterizes high-risk gastric cancer precursors.

Gastric cancer precursors demonstrate highly-variable rates of progression toward neoplasia. Certain high-risk precursors, such as gastric intestinal metaplasia with advanced histologic features, may be at up to 30-fold increased risk for progression compared to lower-risk intestinal metaplasia. The biological differences between high- and low-risk lesions have been incompletely explored.

Single-cell spatial mapping reveals alteration of cell type composition and tissue microenvironment during early colorectal cancer formation.

Colorectal cancer (CRC) is the third leading cause of cancer mortality in the United States. Familial adenomatous polyposis (FAP) is a hereditary syndrome that raises the risk of developing CRC, with total colectomy as the only effective prevention. Even though FAP is rare (0.

Multiomic analysis of familial adenomatous polyposis reveals molecular pathways associated with early tumorigenesis.

Familial adenomatous polyposis (FAP) is a genetic disease causing hundreds of premalignant polyps in affected persons and is an ideal model to study transitions of early precancer states to colorectal cancer (CRC). We performed deep multiomic profiling of 93 samples, including normal mucosa, benign polyps and dysplastic polyps, from six persons with FAP. Transcriptomic, proteomic, metabolomic and lipidomic analyses revealed a dynamic choreography of thousands of molecular and cellular events that occur during precancerous transitions toward cancer formation.

Global loss of promoter-enhancer connectivity and rebalancing of gene expression during early colorectal cancer carcinogenesis.

Although three-dimensional (3D) genome architecture is crucial for gene regulation, its role in disease remains elusive. We traced the evolution and malignant transformation of colorectal cancer (CRC) by generating high-resolution chromatin conformation maps of 33 colon samples spanning different stages of early neoplastic growth in persons with familial adenomatous polyposis (FAP). Our analysis revealed a substantial progressive loss of genome-wide cis-regulatory connectivity at early malignancy stages, correlating with nonlinear gene regulation effects.

A pathologist-AI collaboration framework for enhancing diagnostic accuracies and efficiencies.

In pathology, the deployment of artificial intelligence (AI) in clinical settings is constrained by limitations in data collection and in model transparency and interpretability. Here we describe a digital pathology framework, nuclei.io, that incorporates active learning and human-in-the-loop real-time feedback for the rapid creation of diverse datasets and models.

Prediction of immunochemotherapy response for diffuse large B-cell lymphoma using artificial intelligence digital pathology.

Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous and prevalent subtype of aggressive non-Hodgkin lymphoma that poses diagnostic and prognostic challenges, particularly in predicting drug responsiveness. In this study, we used digital pathology and deep learning to predict responses to immunochemotherapy in patients with DLBCL. We retrospectively collected 251 slide images from 216 DLBCL patients treated with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP), with their immunochemotherapy response labels.

Spatially Segregated Macrophage Populations Predict Distinct Outcomes in Colon Cancer.

Tumor-associated macrophages are transcriptionally heterogeneous, but the spatial distribution and cell interactions that shape macrophage tissue roles remain poorly characterized. Here, we spatially resolve five distinct human macrophage populations in normal and malignant human breast and colon tissue and reveal their cellular associations. This spatial map reveals that distinct macrophage populations reside in spatially segregated micro-environmental niches with conserved cellular compositions that are repeated across healthy and diseased tissue.

Generation of synthetic whole-slide image tiles of tumours from RNA-sequencing data via cascaded diffusion models.

Training machine-learning models with synthetically generated data can alleviate the problem of data scarcity when acquiring diverse and sufficiently large datasets is costly and challenging. Here we show that cascaded diffusion models can be used to synthesize realistic whole-slide image tiles from latent representations of RNA-sequencing data from human tumours. Alterations in gene expression affected the composition of cell types in the generated synthetic image tiles, which accurately preserved the distribution of cell types and maintained the cell fraction observed in bulk RNA-sequencing data, as we show for lung adenocarcinoma, kidney renal papillary cell carcinoma, cervical squamous cell carcinoma, colon adenocarcinoma and glioblastoma.

Inflamed immune phenotype predicts favorable clinical outcomes of immune checkpoint inhibitor therapy across multiple cancer types.

Background: The inflamed immune phenotype (IIP), defined by enrichment of tumor-infiltrating lymphocytes (TILs) within intratumoral areas, is a promising tumor-agnostic biomarker of response to immune checkpoint inhibitor (ICI) therapy. However, it is challenging to define the IIP in an objective and reproducible manner during manual histopathologic examination. Here, we investigate artificial intelligence (AI)-based immune phenotypes capable of predicting ICI clinical outcomes in multiple solid tumor types.

A spatially mapped gene expression signature for intestinal stem-like cells identifies high-risk precursors of gastric cancer.

Objective: Gastric intestinal metaplasia () is a precancerous lesion that increases gastric cancer () risk. The Operative Link on GIM () is a combined clinical-histopathologic system to risk-stratify patients with GIM. The identification of molecular biomarkers that are indicators for advanced OLGIM lesions may improve cancer prevention efforts.

Synthetic whole-slide image tile generation with gene expression profile-infused deep generative models.

In this work, we propose an approach to generate whole-slide image (WSI) tiles by using deep generative models infused with matched gene expression profiles. First, we train a variational autoencoder (VAE) that learns a latent, lower-dimensional representation of multi-tissue gene expression profiles. Then, we use this representation to infuse generative adversarial networks (GANs) that generate lung and brain cortex tissue tiles, resulting in a new model that we call RNA-GAN.

Ampullary Adenocarcinoma, Version 1.2023, NCCN Clinical Practice Guidelines in Oncology.

Ampullary cancers refer to tumors originating from the ampulla of Vater (the ampulla, the intraduodenal portion of the bile duct, and the intraduodenal portion of the pancreatic duct), while periampullary cancers may arise from locations encompassing the head of the pancreas, distal bile duct, duodenum, or ampulla of Vater. Ampullary cancers are rare gastrointestinal malignancies, and prognosis varies greatly based on factors such as patient age, TNM classification, differentiation grade, and treatment modality received. Systemic therapy is used in all stages of ampullary cancer, including neoadjuvant therapy, adjuvant therapy, and first-line or subsequent-line therapy for locally advanced, metastatic, and recurrent disease.

Predicting lymph node metastasis from primary tumor histology and clinicopathologic factors in colorectal cancer using deep learning.

Background: Presence of lymph node metastasis (LNM) influences prognosis and clinical decision-making in colorectal cancer. However, detection of LNM is variable and depends on a number of external factors. Deep learning has shown success in computational pathology, but has struggled to boost performance when combined with known predictors.

Xanthogranulomatous inflammation requiring small bowel anastomosis revision: A case report.

Background: Xanthogranulomatous inflammation (XGI) is an uncommon process involving an accumulation of inflammatory cells, commonly lipid-laden macrophages. XGI has been described to occur throughout the body but only rarely in the lower gastrointestinal tract. We describe a case of XGI contributing to chronic obstructive symptoms in the terminal ileum, in which the patient had an initial diagnostic laparoscopy, continued to have symptoms, then proceeded to have the definitive treatment.

Small bowel adenocarcinoma in neoterminal ileum in setting of stricturing Crohn's disease: A case report and review of literature.

Background: Small bowel adenocarcinomas (SBA) are rare malignancies with exceedingly low survival rates, with different presentation in Crohn's disease (CD). CD-induced SBA poses diagnostic challenges given overlapping presentation with stricturing CD and lack of diagnostics for early detection. Moreover, guidance is lacking on the impact of recently approved therapeutics in CD on SBA management.

PlexusNet: A neural network architectural concept for medical image classification.

State-of-the-art (SOTA) convolutional neural network models have been widely adapted in medical imaging and applied to address different clinical problems. However, the complexity and scale of such models may not be justified in medical imaging and subject to the available resource budget. Further increasing the number of representative feature maps for the classification task decreases the model explainability.

A spatial map of human macrophage niches reveals context-dependent macrophage functions in colon and breast cancer.

Tumor-associated macrophages (TAMs) display heterogeneous phenotypes. Yet the exact tissue cues that shape macrophage functional diversity are incompletely understood. Here we discriminate, spatially resolve and reveal the function of five distinct macrophage niches within malignant and benign breast and colon tissue.

RNA-to-image multi-cancer synthesis using cascaded diffusion models.

Data scarcity presents a significant obstacle in the field of biomedicine, where acquiring diverse and sufficient datasets can be costly and challenging. Synthetic data generation offers a potential solution to this problem by expanding dataset sizes, thereby enabling the training of more robust and generalizable machine learning models. Although previous studies have explored synthetic data generation for cancer diagnosis, they have predominantly focused on single modality settings, such as whole-slide image tiles or RNA-Seq data.

Development and Validation of a Machine Learning Model for Detection and Classification of Tertiary Lymphoid Structures in Gastrointestinal Cancers.

Importance: Tertiary lymphoid structures (TLSs) are associated with a favorable prognosis and improved response to cancer immunotherapy. The current approach for evaluation of TLSs is limited by interobserver variability and high complexity and cost of specialized imaging techniques.

Objective: To develop a machine learning model for automated and quantitative evaluation of TLSs based on routine histopathology images.