Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Prototype Instance-semantic Disentanglement with Low-rank Regularized Subspace Clustering for WSIs Explainable Recognition

Chentao Li, Pan Huang

TL;DR

This work tackles instance-semantic entanglement in WSI MIL by introducing PID-LRSC, which fuses two components: (i) a two-phase low-rank regularized subspace clustering to separate tumor, non-tumor, and background subspaces under severe class imbalance, and (ii) prototype-based instance semantic disentanglement guided by a distributional distance (CDF) to align subspaces with tumor prototypes. The framework is trained end-to-end with a loss that jointly optimizes classification and subspace separation, improving both accuracy and interpretability. Extensive experiments on multicentre pathology datasets show state-of-the-art performance and clearer, more reliable decision evidence, enhancing the clinical value of WSI-based diagnosis.

Abstract

The tumor region plays a key role in pathological diagnosis. Tumor tissues are highly similar to precancerous lesions and non tumor instances often greatly exceed tumor instances in whole slide images (WSIs). These issues cause instance-semantic entanglement in multi-instance learning frameworks, degrading both model representation capability and interpretability. To address this, we propose an end-to-end prototype instance semantic disentanglement framework with low-rank regularized subspace clustering, PID-LRSC, in two aspects. First, we use secondary instance subspace learning to construct low-rank regularized subspace clustering (LRSC), addressing instance entanglement caused by an excessive proportion of non tumor instances. Second, we employ enhanced contrastive learning to design prototype instance semantic disentanglement (PID), resolving semantic entanglement caused by the high similarity between tumor and precancerous tissues. We conduct extensive experiments on multicentre pathology datasets, implying that PID-LRSC outperforms other SOTA methods. Overall, PID-LRSC provides clearer instance semantics during decision-making and significantly enhances the reliability of auxiliary diagnostic outcomes.

Prototype Instance-semantic Disentanglement with Low-rank Regularized Subspace Clustering for WSIs Explainable Recognition

TL;DR

This work tackles instance-semantic entanglement in WSI MIL by introducing PID-LRSC, which fuses two components: (i) a two-phase low-rank regularized subspace clustering to separate tumor, non-tumor, and background subspaces under severe class imbalance, and (ii) prototype-based instance semantic disentanglement guided by a distributional distance (CDF) to align subspaces with tumor prototypes. The framework is trained end-to-end with a loss that jointly optimizes classification and subspace separation, improving both accuracy and interpretability. Extensive experiments on multicentre pathology datasets show state-of-the-art performance and clearer, more reliable decision evidence, enhancing the clinical value of WSI-based diagnosis.

Abstract

The tumor region plays a key role in pathological diagnosis. Tumor tissues are highly similar to precancerous lesions and non tumor instances often greatly exceed tumor instances in whole slide images (WSIs). These issues cause instance-semantic entanglement in multi-instance learning frameworks, degrading both model representation capability and interpretability. To address this, we propose an end-to-end prototype instance semantic disentanglement framework with low-rank regularized subspace clustering, PID-LRSC, in two aspects. First, we use secondary instance subspace learning to construct low-rank regularized subspace clustering (LRSC), addressing instance entanglement caused by an excessive proportion of non tumor instances. Second, we employ enhanced contrastive learning to design prototype instance semantic disentanglement (PID), resolving semantic entanglement caused by the high similarity between tumor and precancerous tissues. We conduct extensive experiments on multicentre pathology datasets, implying that PID-LRSC outperforms other SOTA methods. Overall, PID-LRSC provides clearer instance semantics during decision-making and significantly enhances the reliability of auxiliary diagnostic outcomes.
Paper Structure (12 sections, 8 equations, 4 figures, 3 tables)

This paper contains 12 sections, 8 equations, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Problem Statement
  4. Two-phase Coarse-to-fine LRSC
  5. Latent Characteristic function discrepancy based PID
  6. End-to-end Optimization
  7. Experimental Results
  8. Datasets and Setup
  9. Comparison with SOTA Methods
  10. Ablation Experiments
  11. Visualization of WSI Interpretability and Representation
  12. Conclusion

Figures (4)

  • Figure 1: The motivation of proposed PID-LRSC. Left: Instance entanglement caused by excessive proportion of non-tumor over tumor. Right: Semantic entanglement caused by high similarity between precancerous lesions and tumor.
  • Figure 2: Overview of PID-LRSC. Cropped instances extracted by encoder are clustered into three subspaces by Low Rank Ssubspace Clustering method in two phases. They are then identified into tumor, non-tumor and background by Prototype Instance-semantic Disentanglement. End-to-end optimization is performed for disentangled representation learning with refined instance features.
  • Figure 3: Visualization results on multicentre datasets. The first column is original WSI with pathologist-annotated boundaries. The second is the attention contribution scores corresponding to its category. The last column is the disentangling results.
  • Figure 4: ANOVA plots on four datasets where $\eta_*$ denote the value of effect size. Statistically, $\eta_*>0.4$ implies the model having large effects within categories.