Compact and De-Biased Negative Instance Embedding for Multi-Instance Learning on Whole-Slide Image Classification

2024-02-16

IEEE International Conference on Acoustics, Speech, and Signal Processing

AI Summary

In whole-slide image (WSI) classification, multi-instance learning (MIL) suffers from poor generalization due to unlabeled patches and inter-slide staining variations. To address this, we propose a semi-supervised MIL framework that leverages the inherent negative supervision signal—i.e., all patches in normal WSIs are naturally negative—to construct compact, bias-mitigated negative embeddings, enabling orthogonal bias correction relative to the MIL backbone. Our method explicitly models common intra-class variations among normal samples by integrating stain-invariant feature extraction, contrastive learning, and consistency regularization. Evaluated on Camelyon-16 and TCGA lung cancer datasets, our approach significantly outperforms state-of-the-art MIL and existing semi-supervised methods, yielding improved robustness and generalization across staining conditions. The source code is publicly available.

Technology Category

• Machine Learning: Semi-Supervised Learning
• Computer Vision: Multi-modal Vision
• Search and Optimization: Learning to Search

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: Graph embeddings and representation learning for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Whole-slide image (WSI) classification is a challenging task because 1) patches from WSI lack annotation, and 2) WSI possesses unnecessary variability, e.g., stain protocol. Recently, Multiple-Instance Learning (MIL) has made significant progress, allowing for classification based on slide-level, rather than patch-level, annotations. However, existing MIL methods ignore that all patches from normal slides are normal. Using this free annotation, we introduce a semi-supervision signal to de-bias the inter-slide variability and to capture the common factors of variation within normal patches. Because our method is orthogonal to the MIL algorithm, we evaluate our method on top of the recently proposed MIL algorithms and also compare the performance with other semi-supervised approaches. We evaluate our method on two public WSI datasets including Camelyon-16 and TCGA lung cancer and demonstrate that our approach significantly improves the predictive performance of existing MIL algorithms and outperforms other semi-supervised algorithms. We release our code at https://github.com/AITRICS/pathology_mil.

Problem

• De-biasing inter-slide variability in WSI classification
• Leveraging normal patches as free annotation source
• Improving MIL performance through semi-supervision signals

Innovation

• Semi-supervision signal for de-biasing variability
• Leveraging free annotation from normal patches
• Orthogonal integration with existing MIL algorithms