M3d-CAM: A PyTorch library to generate 3D data attention maps for medical deep learning

TL;DR

M3d-CAM is a user-friendly PyTorch library that generates interpretable attention maps for 2D and 3D medical data, supporting multiple visualization methods with minimal code effort.

Contribution

It introduces an easy-to-use, versatile library for generating attention maps in medical deep learning models, supporting various methods and data types with minimal implementation effort.

Findings

Supports multiple attention map methods like Guided Backpropagation and Grad-CAM

Works seamlessly with 2D and 3D medical data for classification and segmentation

Requires only a single line of code in most cases

Abstract

M3d-CAM is an easy to use library for generating attention maps of CNN-based PyTorch models improving the interpretability of model predictions for humans. The attention maps can be generated with multiple methods like Guided Backpropagation, Grad-CAM, Guided Grad-CAM and Grad-CAM++. These attention maps visualize the regions in the input data that influenced the model prediction the most at a certain layer. Furthermore, M3d-CAM supports 2D and 3D data for the task of classification as well as for segmentation. A key feature is also that in most cases only a single line of code is required for generating attention maps for a model making M3d-CAM basically plug and play.