mmNormVAE: Normative Modeling on Multimodal Neuroimaging Data using Variational Autoencoders

Sayantan Kumar, Philip Payne, Aristeidis Sotiras

Keywords: normative modelling, multimodal variational autoencoder, Product-of-Experts, Alzheimer's Disease

TL;DR: We designed a multi-modal normative modelling framework based on multimodal variational autoencoders (mmNormVAE) where disease (AD) abnormality is aggregated across multiple neuroimaging modalities.

Abstract: Normative modelling is a popular method for studying brain disorders like Alzheimer's Disease (AD) where the normal brain patterns of cognitively normal subjects are modelled and can be used at subject-level to detect deviations relating to disease pathology. So far, deep learning-based normative frameworks have largely been applied on a single imaging modality. We aim to design a multi-modal normative modelling framework based on multimodal variational autoencoders (mmNormVAE) where disease abnormality is aggregated across multiple neuroimaging modalities (T1-weighted and T2-weighted MRI) and subsequently used to estimate subject-level neuroanatomical deviations due to AD.

Submission Number: 54