SI-MIL: Taming Deep MIL for Self-Interpretability in Gigapixel Histopathology

TL;DR

SI-MIL introduces an inherently interpretable deep learning framework for gigapixel histopathology images, providing feature-level explanations and maintaining competitive predictive performance across multiple cancer types.

Contribution

The paper presents SI-MIL, a novel self-interpretable MIL method that combines deep learning with handcrafted pathological features for transparent WSI analysis.

Findings

SI-MIL achieves competitive accuracy on WSI-level prediction tasks.

It provides feature-level interpretability rooted in pathological insights.

Benchmarking shows SI-MIL's interpretability is user-friendly and faithful.

Abstract

Introducing interpretability and reasoning into Multiple Instance Learning (MIL) methods for Whole Slide Image (WSI) analysis is challenging, given the complexity of gigapixel slides. Traditionally, MIL interpretability is limited to identifying salient regions deemed pertinent for downstream tasks, offering little insight to the end-user (pathologist) regarding the rationale behind these selections. To address this, we propose Self-Interpretable MIL (SI-MIL), a method intrinsically designed for interpretability from the very outset. SI-MIL employs a deep MIL framework to guide an interpretable branch grounded on handcrafted pathological features, facilitating linear predictions. Beyond identifying salient regions, SI-MIL uniquely provides feature-level interpretations rooted in pathological insights for WSIs. Notably, SI-MIL, with its linear prediction constraints, challenges the prevalent myth of an inevitable trade-off between model interpretability and performance, demonstrating competitive results compared to state-of-the-art methods on WSI-level prediction tasks across three cancer types. In addition, we thoroughly benchmark the local and global-interpretability of SI-MIL in terms of statistical analysis, a domain expert study, and desiderata of interpretability, namely, user-friendliness and faithfulness.