xMIL: Insightful Explanations for Multiple Instance Learning in Histopathology
This work addresses the need for better explainable AI in digital histopathology, enabling pathologists to gain insights and debug models, though it is incremental as it builds on existing MIL and XAI techniques.
The paper tackled the problem of limited explanation methods for multiple instance learning (MIL) in histopathology, which lag behind in handling large bag sizes and instance interactions, and introduced xMIL, a refined framework that outperforms previous methods with improved faithfulness scores on biomarker prediction tasks.
Multiple instance learning (MIL) is an effective and widely used approach for weakly supervised machine learning. In histopathology, MIL models have achieved remarkable success in tasks like tumor detection, biomarker prediction, and outcome prognostication. However, MIL explanation methods are still lagging behind, as they are limited to small bag sizes or disregard instance interactions. We revisit MIL through the lens of explainable AI (XAI) and introduce xMIL, a refined framework with more general assumptions. We demonstrate how to obtain improved MIL explanations using layer-wise relevance propagation (LRP) and conduct extensive evaluation experiments on three toy settings and four real-world histopathology datasets. Our approach consistently outperforms previous explanation attempts with particularly improved faithfulness scores on challenging biomarker prediction tasks. Finally, we showcase how xMIL explanations enable pathologists to extract insights from MIL models, representing a significant advance for knowledge discovery and model debugging in digital histopathology. Codes are available at: https://github.com/bifold-pathomics/xMIL.