MacNet: An End-to-End Manifold-Constrained Adaptive Clustering Network for Interpretable Whole Slide Image Classification

TL;DR

This work proposes an end-to-end MIL framework that integrates Grassmann re-embedding and manifold adaptive clustering, where the manifold geometric structure facilitates robust clustering results and designs a prior knowledge guiding proxy instance labeling and aggregation strategy to approximate patch labels and focus on pathologically relevant tumor regions.

Abstract

Whole slide images (WSIs) are the gold standard for pathological diagnosis and sub-typing. Current main-stream two-step frameworks employ offline feature encoders trained without domain-specific knowledge. Among them, attention-based multiple instance learning (MIL) methods are outcome-oriented and offer limited interpretability. Clustering-based approaches can provide explainable decision-making process but suffer from high dimension features and semantically ambiguous centroids. To this end, we propose an end-to-end MIL framework that integrates Grassmann re-embedding and manifold adaptive clustering, where the manifold geometric structure facilitates robust clustering results. Furthermore, we design a prior knowledge guiding proxy instance labeling and aggregation strategy to approximate patch labels and focus on pathologically relevant tumor regions. Experiments on multicentre WSI datasets demonstrate that: 1) our cluster-incorporated model achieves superior performance in both grading accuracy and interpretability; 2) end-to-end learning refines better feature representations and it requires acceptable computation resources.

Figures (11)

• Figure 1: Motivation of MacNet. a): Attention heatmap only reflects outcome-guided scores. b): Traditional clustering suffers from high dimension and ambiguous centroids. c): MacNet obtains robust clustering results and transparent decision-making.
• Figure 2: Overall flowchart of proposed end-to-end MacNet. Prior knowledge instances are annotated by pathologists.
• Figure 3: a): Geodesic distance calculated by discrete principal angles. b): Grassmann Re-embedding maps features into infinite dimensional space.
• Figure 4: Confusion Matrix plots on multicentre datasets.
• Figure 5: Visualization results on AMU-LSCC and DHMC-LUNG. The first column is original WSI. The second column is labeling result during prediction. The last column is the attention heatmap corresponding to its category.