ProtoMIL: Multiple Instance Learning with Prototypical Parts for Whole-Slide Image Classification
It addresses the problem of interpretability in MIL for applications like medical imaging, offering a novel approach that combines accuracy with fine-grained explanations.
The paper tackles the lack of explainability in Multiple Instance Learning (MIL) by introducing ProtoMIL, a self-explainable method that uses visual prototypes, achieving competitive accuracy on five recognized datasets.
Multiple Instance Learning (MIL) gains popularity in many real-life machine learning applications due to its weakly supervised nature. However, the corresponding effort on explaining MIL lags behind, and it is usually limited to presenting instances of a bag that are crucial for a particular prediction. In this paper, we fill this gap by introducing ProtoMIL, a novel self-explainable MIL method inspired by the case-based reasoning process that operates on visual prototypes. Thanks to incorporating prototypical features into objects description, ProtoMIL unprecedentedly joins the model accuracy and fine-grained interpretability, which we present with the experiments on five recognized MIL datasets.