Brain Structural Saliency Over The Ages

Daniel Petrus Taylor, Jonathan Phillip Shock, Deshendran Moodley, Jonathan Ipser, Matthias S. Treder

Keywords: Brain Age, MRI, Layer-wise Relevance Propagation, DeepLIFT

TL;DR: Relevance for Brain Age is distributed according to region-specific functions of chronological age, and trends in relevance distribution are largely consistent between attribution methods.

Abstract:

Brain Age (BA) estimation via Deep Learning has become a strong and reliable bio-marker for brain health, but the black-box nature of Neural Networks does not easily allow insight into the causal features of brain ageing. We trained a ResNet model as a BA regressor on T1 structural MRI volumes from a small cross-sectional cohort of $524$ individuals. Using Layer-wise Relevance Propagation (LRP) and DeepLIFT saliency mapping techniques, we analyse the trained model to determine the most revealing structures over the course of brain ageing for a Deep Convolutional Network, and compare these between the saliency mapping techniques.

We show the change in attribution of relevance to different brain regions through the course of ageing, and which regions show the least and greatest change over time. We also examine the effect of Brain Age Delta (DBA) on the distribution of relevance within the brain volume.

A tripartite pattern of relevance attribution to brain regions emerges. Some regions increase in relevance with age (e.g. the right Transverse Temporal Gyrus, part of the auditory cortex known to be affected by healthy ageing); some decrease in relevance with age (e.g. the right Fourth Ventricle, known to dilate with age); and others remained consistently relevant across ages.

It is hoped that these findings will provide clinically relevant region-wise trajectories for normal brain ageing, and a baseline against which to compare brain ageing trajectories.

Registration: I acknowledge that publication of this at MIDL and in the proceedings requires at least one of the authors to register and present the work during the conference.

Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.

Paper Type: validation/application paper

Primary Subject Area: Interpretability and Explainable AI

Secondary Subject Area: Detection and Diagnosis

Confidentiality And Author Instructions: I read the call for papers and author instructions. I acknowledge that exceeding the page limit and/or altering the latex template can result in desk rejection.