Interpretation of 3D CNNs for Brain MRI Data Classification

TL;DR

This paper explores voxel-wise interpretation of 3D CNNs applied to brain MRI data for gender classification, comparing three interpretation methods and providing an open-source library.

Contribution

It introduces a voxel-wise interpretation framework for 3D CNNs on brain MRI, extending previous region-based analyses and offering an open-source tool.

Findings

Voxel-wise interpretation reveals detailed brain differences.

Comparison of three interpretation methods shows their relative strengths.

Open-source library facilitates further research in medical image analysis.

Abstract

Deep learning shows high potential for many medical image analysis tasks. Neural networks can work with full-size data without extensive preprocessing and feature generation and, thus, information loss. Recent work has shown that the morphological difference in specific brain regions can be found on MRI with the means of Convolution Neural Networks (CNN). However, interpretation of the existing models is based on a region of interest and can not be extended to voxel-wise image interpretation on a whole image. In the current work, we consider the classification task on a large-scale open-source dataset of young healthy subjects -- an exploration of brain differences between men and women. In this paper, we extend the previous findings in gender differences from diffusion-tensor imaging on T1 brain MRI scans. We provide the voxel-wise 3D CNN interpretation comparing the results of three interpretation methods: Meaningful Perturbations, Grad CAM and Guided Backpropagation, and contribute with the open-source library.