Work

A lightweight generative model for interpretable subject-level prediction

Year: 2024

Type: article

Abstract: Recent years have seen a growing interest in methods for predicting an unknown variable of interest, such as a subject’s diagnosis, from medical images depicting its anatomical-functional effects. Methods based on discriminative modeling excel at making accurate predictions, but are challenged in their ability to explain their decisions in anatomically meaningful terms. In this paper, we propose a simple technique for single-subject prediction that is inherently interpretable. It augments the generative models used in classical human brain mapping techniques, in which the underlying cause–effect relations can be encoded, with a multivariate noise model that captures dominant spatial correlations. Experiments demonstrate that the resulting model can be efficiently inverted to make accurate subject-level predictions, while at the same time offering intuitive visual explanations of its inner workings. The method is easy to use: training is fast for typical training set sizes, and only a single hyperparameter needs to be set by the user. Our code is available at https://github.com/chiara-mauri/Interpretable-subject-level-prediction . • We propose an inherently interpretable generative model for image-based prediction. • The method is also accurate, easy to use, and fast to train. • Besides predictions, it offers intuitive causal explanations of its inner workings. • We tested the method on brain age and gender prediction tasks on UK Biobank data.

Source: Medical Image Analysis

DOI: 10.1016/j.media.2024.103436

Authors: Chiara Mauri, Stefano Cerri, Oula Puonti, Mark Mühlau, Koen Van Leemput