Interpretable breast cancer classification using CNNs on mammographic images

TL;DR

This paper evaluates interpretability techniques for CNNs in mammogram classification, highlighting Grad-CAM's effectiveness in revealing tissue patterns and aiding clinical understanding.

Contribution

It compares post-hoc interpretability methods for CNNs in breast cancer detection, emphasizing Grad-CAM's superior explanatory insights on mammographic images.

Findings

Grad-CAM provides detailed explanations of CNN decisions.

Grad-CAM reveals tissue pattern differences between normal, benign, and malignant cases.

Interpretability techniques can enhance clinical trust in AI models.

Abstract

Deep learning models have achieved promising results in breast cancer classification, yet their 'black-box' nature raises interpretability concerns. This research addresses the crucial need to gain insights into the decision-making process of convolutional neural networks (CNNs) for mammogram classification, specifically focusing on the underlying reasons for the CNN's predictions of breast cancer. For CNNs trained on the Mammographic Image Analysis Society (MIAS) dataset, we compared the post-hoc interpretability techniques LIME, Grad-CAM, and Kernel SHAP in terms of explanatory depth and computational efficiency. The results of this analysis indicate that Grad-CAM, in particular, provides comprehensive insights into the behavior of the CNN, revealing distinctive patterns in normal, benign, and malignant breast tissue. We discuss the implications of the current findings for the use of machine learning models and interpretation techniques in clinical practice.