Are Explanations Helpful? A Comparative Analysis of Explainability Methods in Skin Lesion Classifiers

TL;DR

This paper compares seven explainability methods for skin lesion classifiers, highlighting their ability to reveal biases and the need for more comprehensive explanations for clinical trust.

Contribution

It provides a systematic analysis of existing explainability techniques in skin cancer models, identifying strengths and limitations.

Findings

Explainability methods reveal biases in skin-lesion models.

Current techniques lack comprehensiveness for full transparency.

Room for improvement in explanation quality for clinical use.

Abstract

Deep Learning has shown outstanding results in computer vision tasks; healthcare is no exception. However, there is no straightforward way to expose the decision-making process of DL models. Good accuracy is not enough for skin cancer predictions. Understanding the model's behavior is crucial for clinical application and reliable outcomes. In this work, we identify desiderata for explanations in skin-lesion models. We analyzed seven methods, four based on pixel-attribution (Grad-CAM, Score-CAM, LIME, SHAP) and three on high-level concepts (ACE, ICE, CME), for a deep neural network trained on the International Skin Imaging Collaboration Archive. Our findings indicate that while these techniques reveal biases, there is room for improving the comprehensiveness of explanations to achieve transparency in skin-lesion models.