Model Agnostic Saliency for Weakly Supervised Lesion Detection from Breast DCE-MRI

TL;DR

This paper introduces MASD, a model-agnostic saliency method tailored for weakly supervised breast lesion detection in DCE-MRI, addressing key interpretability challenges and achieving state-of-the-art results.

Contribution

The paper proposes a novel, model-agnostic 1-class saliency detector that satisfies specific lesion detection interpretability conditions in medical imaging.

Findings

Achieves state-of-the-art detection accuracy

Addresses interpretability issues in lesion visualization

Outperforms existing visualization methods

Abstract

There is a heated debate on how to interpret the decisions provided by deep learning models (DLM), where the main approaches rely on the visualization of salient regions to interpret the DLM classification process. However, these approaches generally fail to satisfy three conditions for the problem of lesion detection from medical images: 1) for images with lesions, all salient regions should represent lesions, 2) for images containing no lesions, no salient region should be produced,and 3) lesions are generally small with relatively smooth borders. We propose a new model-agnostic paradigm to interpret DLM classification decisions supported by a novel definition of saliency that incorporates the conditions above. Our model-agnostic 1-class saliency detector (MASD) is tested on weakly supervised breast lesion detection from DCE-MRI, achieving state-of-the-art detection accuracy when compared to current visualization methods.