GAS-MIL: Group-Aggregative Selection Multi-Instance Learning for Ensemble of Foundation Models in Digital Pathology Image Analysis

2025/10/03 by Peiran Quan, Zifan Gu, Quan, Peiran +13
Computer Science · #AI in cancer detection #Artificial Intelligence (cs.AI) #Computer Vision and Pattern Recognition (cs.CV) #Digital Imaging for Blood Diseases #FOS: Computer and information sciences

paper · pdf · doi:10.48550/arxiv.2510.03555

openalex publication_date 2025/10/03 · openalex created_date 2025/10/09 · openalex updated_date 2026/07/28

Abstract

Foundation models (FMs) have transformed computational pathology by providing powerful, general-purpose feature extractors. However, adapting and benchmarking individual FMs for specific diagnostic tasks is often time-consuming and resource-intensive, especially given their scale and diversity. To address this challenge, we introduce Group-Aggregative Selection Multi-Instance Learning (GAS-MIL), a flexible ensemble framework that seamlessly integrates features from multiple FMs, preserving their complementary strengths without requiring manual feature selection or extensive task-specific fine-tuning. Across classification tasks in three cancer datasets-prostate (PANDA), ovarian (UBC-OCEAN), and breast (TCGA-BrCa)-GAS-MIL consistently achieves superior or on-par performance relative to individual FMs and established MIL methods, demonstrating its robustness and generalizability. By enabling efficient integration of heterogeneous FMs, GAS-MIL streamlines model deployment for pathology and provides a scalable foundation for future multimodal and precision oncology applications.

Citations

• Benchmarking Pathology Foundation Models: Adaptation Strategies and Scenarios
• Phikon-v2, A large and public feature extractor for biomarker prediction
• Advances in Multiple Instance Learning for Whole Slide Image Analysis: Techniques, Challenges, and Future Directions
• LongNet: Scaling Transformers to 1,000,000,000 Tokens
• LVM-Med: Learning Large-Scale Self-Supervised Vision Models for Medical Imaging via Second-order Graph Matching
• DINOv2: Learning Robust Visual Features without Supervision
• Micrograph segmentations for DDEVD
• Segment Anything
• iBOT: Image BERT Pre-Training with Online Tokenizer
• Masked Autoencoders Are Scalable Vision Learners
• TransMIL: Transformer based Correlated Multiple Instance Learning for Whole Slide Image Classification
• Emerging Properties in Self-Supervised Vision Transformers
• Learning Transferable Visual Models From Natural Language Supervision
• Language Models are Few-Shot Learners
• The 2019 International Society of Urological Pathology (ISUP) Consensus Conference on Grading of Prostatic Carcinoma
• A Generalized Deep Learning Framework for Whole-Slide Image Segmentation and Analysis
• Multimodal Multitask Representation Learning for Pathology Biobank Metadata Prediction
• EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
• Attention-based Deep Multiple Instance Learning
• Classification and Disease Localization in Histopathology Using Only Global Labels: A Weakly-Supervised Approach
• Deep Residual Learning for Image Recognition
• A Threshold Selection Method from Gray-Level Histograms

Related

• Benchmarking histopathology foundation models in a multi-center dataset for skin cancer subtyping
• nnMIL: A generalizable multiple instance learning framework for computational pathology
• Ensemble learning of foundation models for precision oncology
• Towards Large-Scale Training of Pathology Foundation Models
• Foundation Models for Slide-level Cancer Subtyping in Digital Pathology