Contrastive Cross-Bag Augmentation for Multiple Instance Learning-based Whole Slide Image Classification

2025-08-05

AI Summary

In multi-instance learning (MIL) for whole-slide image (WSI) classification, insufficient pseudo-bag diversity hampers model generalization. To address this, we propose a cross-bag contrastive augmentation framework: first, a class-consistent cross-bag instance sampling strategy dynamically constructs high-diversity pseudo-bags from all bags of the same class; second, a dual-level (bag-level and group-level) contrastive learning scheme is introduced to enable fine-grained semantic discrimination and feature alignment. This design mitigates model bias induced by over-amplification of critical instances, particularly enhancing detection of sparse, small tumor regions. Extensive experiments on multiple public WSI benchmarks demonstrate that our method consistently outperforms state-of-the-art approaches, achieving significant improvements in classification accuracy and robustness.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Multi-modal Vision
• Multiagent Systems: Multiagent Learning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Recent pseudo-bag augmentation methods for Multiple Instance Learning (MIL)-based Whole Slide Image (WSI) classification sample instances from a limited number of bags, resulting in constrained diversity. To address this issue, we propose Contrastive Cross-Bag Augmentation ($C^2Aug$) to sample instances from all bags with the same class to increase the diversity of pseudo-bags. However, introducing new instances into the pseudo-bag increases the number of critical instances (e.g., tumor instances). This increase results in a reduced occurrence of pseudo-bags containing few critical instances, thereby limiting model performance, particularly on test slides with small tumor areas. To address this, we introduce a bag-level and group-level contrastive learning framework to enhance the discrimination of features with distinct semantic meanings, thereby improving model performance. Experimental results demonstrate that $C^2Aug$ consistently outperforms state-of-the-art approaches across multiple evaluation metrics.

Problem

• Enhancing pseudo-bag diversity in MIL-based WSI classification
• Reducing critical instance imbalance in augmented pseudo-bags
• Improving feature discrimination for small tumor area detection

Innovation

• Contrastive Cross-Bag Augmentation for diversity
• Bag-level and group-level contrastive learning
• Enhances discrimination of semantic features