Molecular-driven Foundation Model for Oncologic Pathology
read the original abstract
Foundation models are reshaping computational pathology by enabling transfer learning, where models pre-trained on vast datasets can be adapted for downstream diagnostic, prognostic, and therapeutic response tasks. Despite these advances, foundation models are still limited in their ability to encode the entire gigapixel whole-slide images without additional training and often lack complementary multimodal data. Here, we introduce Threads, a slide-level foundation model capable of generating universal representations of whole-slide images of any size. Threads was pre-trained using a multimodal learning approach on a diverse cohort of 47,171 hematoxylin and eosin (H&E)-stained tissue sections, paired with corresponding genomic and transcriptomic profiles - the largest such paired dataset to be used for foundation model development to date. This unique training paradigm enables Threads to capture the tissue's underlying molecular composition, yielding powerful representations applicable to a wide array of downstream tasks. In extensive benchmarking across 54 oncology tasks, including clinical subtyping, grading, mutation prediction, immunohistochemistry status determination, treatment response prediction, and survival prediction, Threads outperformed all baselines while demonstrating remarkable generalizability and label efficiency. It is particularly well suited for predicting rare events, further emphasizing its clinical utility. We intend to make the model publicly available for the broader community.
This paper has not been read by Pith yet.
Forward citations
Cited by 9 Pith papers
-
Data-Efficient Multimodal Alignment for Histopathology-based Molecular Prediction
Lightweight contrastive alignment of frozen histopathology and RNA-seq foundation models enables open-vocabulary molecular prediction from H&E slides, reporting 25-fold retrieval gains on a 1720-sample multi-cancer co...
-
Democratizing and accelerating AI-driven pathology research through agentic intelligence
PathLab is an agentic framework that translates natural-language objectives into validated computational pathology workflows, achieving non-inferior performance on 12 datasets across four task families while enabling ...
-
Clinically-Informed Modeling for Pediatric Brain Tumor Classification from Whole-Slide Histopathology Images
An expert-guided contrastive fine-tuning framework improves fine-grained slide-level classification of pediatric brain tumors under low-data and class-imbalanced conditions by regularizing representations with clinica...
-
Mitigating Batch Effects in Histopathology via Language-Mediated Robust Embedding Generation
GLMP generates robust pathology embeddings by routing histology images through an intermediate textual representation produced by general-purpose MLLMs to mitigate batch effects.
-
Predicting Immune Biomarkers with MultiModal Mixture-of-Expert Pathology Foundation Models Empowers Precision Oncology
MixTIME uses a learnable-router MoE to fuse three pathology foundation models for pixel- and slide-level prediction of 17 mIF protein markers from H&E images, improving spatial domain ID, survival prediction, and path...
-
AGE-MIL: Anchor-Guided Evidence Learning for Patient-Level Prediction
AGE-MIL is a new MIL framework that uses patient-level anchors to direct patch selection and evidence accumulation for stable patient-level predictions on six pathology tasks, outperforming eight prior MIL methods.
-
Isotonic Survival Regression: Calibrated Survival Distributions from Deep Cox Models
Post-hoc isotonic regression calibration for deep Cox survival models that improves calibration with theoretical guarantees including double-robustness and asymptotic calibration.
-
Clinically-Informed Modeling for Pediatric Brain Tumor Classification from Whole-Slide Histopathology Images
An expert-guided contrastive fine-tuning framework improves fine-grained slide-level classification of pediatric brain tumors under low-sample and class-imbalanced conditions.
-
Genetically Aligned Patient Representations Improve Hematological Diagnosis
A two-stage training method using self-supervised pretraining on cell images followed by contrastive alignment with genetic data creates improved patient encoders for hematological diagnosis.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.