Multiple Instance Learning with Coarse-to-Fine Self-Distillation

2025-02-04

AI Summary

In computational pathology, weak instance-level representations and insufficient contextual modeling in whole-slide image (WSI) multiple-instance learning (MIL) arise from reliance solely on bag-level supervision. To address this, we propose a Coarse-to-Fine Self-Distillation (CFSD) paradigm that, for the first time, automatically transforms coarse bag-level supervision into high-quality instance-level supervision. We further introduce 2D positional encoding (2DPE) to explicitly model spatial relationships among instances. Theoretically, CFSD guarantees instance learnability. Evaluated on three benchmarks—TCGA-NSCLC, CAMELYON16, and breast cancer receptor status prediction—our method achieves state-of-the-art performance: AUCs of 0.9152 and 0.8524 for estrogen and progesterone receptor prediction, 0.9618 for histological subtype classification, and 0.8634 for tumor detection.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Representation Learning for Vision
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Multiple Instance Learning (MIL) for whole slide image (WSI) analysis in computational pathology often neglects instance-level learning as supervision is typically provided only at the bag level. In this work, we present PathMIL, a framework designed to improve MIL through two perspectives: (1) employing instance-level supervision and (2) learning inter-instance contextual information on bag level. Firstly, we propose a novel Coarse-to-Fine Self-Distillation (CFSD) paradigm, to probe and distil a classifier trained with bag-level information to obtain instance-level labels which could effectively provide the supervision for the same classifier in a finer way. Secondly, to capture inter-instance contextual information in WSI, we propose Two-Dimensional Positional Encoding (2DPE), which encodes the spatial appearance of instances within a bag. We also theoretically and empirically prove the instance-level learnability of CFSD. PathMIL is evaluated on multiple benchmarking tasks, including subtype classification (TCGA-NSCLC), tumour classification (CAMELYON16), and an internal benchmark for breast cancer receptor status classification. Our method achieves state-of-the-art performance, with AUC scores of 0.9152 and 0.8524 for estrogen and progesterone receptor status classification, respectively, an AUC of 0.9618 for subtype classification, and 0.8634 for tumour classification, surpassing existing methods.

Problem

• Improves Multiple Instance Learning in pathology
• Introduces Coarse-to-Fine Self-Distillation paradigm
• Encodes spatial context in whole slide images

Innovation

• Coarse-to-Fine Self-Distillation
• Two-Dimensional Positional Encoding
• Instance-level supervision