Visual Explanations From Deep 3D Convolutional Neural Networks for Alzheimer's Disease Classification
This work addresses the need for interpretability in deep learning models for Alzheimer's disease diagnosis, which is incremental as it builds on existing 3D-CNN methods by adding visualization techniques.
The authors tackled the problem of generating visual explanations from 3D convolutional neural networks for Alzheimer's disease classification, developing three efficient approaches that identify important brain parts for diagnosis, with visual checks and a quantitative localization benchmark indicating their effectiveness.
We develop three efficient approaches for generating visual explanations from 3D convolutional neural networks (3D-CNNs) for Alzheimer's disease classification. One approach conducts sensitivity analysis on hierarchical 3D image segmentation, and the other two visualize network activations on a spatial map. Visual checks and a quantitative localization benchmark indicate that all approaches identify important brain parts for Alzheimer's disease diagnosis. Comparative analysis show that the sensitivity analysis based approach has difficulty handling loosely distributed cerebral cortex, and approaches based on visualization of activations are constrained by the resolution of the convolutional layer. The complementarity of these methods improves the understanding of 3D-CNNs in Alzheimer's disease classification from different perspectives.