Auto-encoding brain networks with applications to analyzing large-scale brain imaging datasets

TL;DR

This paper introduces GATE, a deep learning-based nonlinear latent factor model for analyzing large-scale brain networks, improving prediction and inference of brain-cognition relationships over existing methods.

Contribution

The paper presents GATE, a novel deep learning framework for modeling brain connectomes as networks, capturing complex structures and associations with traits more effectively than prior approaches.

Findings

GATE outperforms competitors in prediction accuracy.

Structural connectomes are strongly linked to cognitive traits.

GATE offers efficient computation and statistical inference.

Abstract

There has been huge interest in studying human brain connectomes inferred from different imaging modalities and exploring their relationship with human traits, such as cognition. Brain connectomes are usually represented as networks, with nodes corresponding to different regions of interest (ROIs) and edges to connection strengths between ROIs. Due to the high-dimensionality and non-Euclidean nature of networks, it is challenging to depict their population distribution and relate them to human traits. Current approaches focus on summarizing the network using either pre-specified topological features or principal components analysis (PCA). In this paper, building on recent advances in deep learning, we develop a nonlinear latent factor model to characterize the population distribution of brain graphs and infer the relationships between brain structural connectomes and human traits. We refer to our method as Graph AuTo-Encoding (GATE). We applied GATE to two large-scale brain imaging datasets, the Adolescent Brain Cognitive Development (ABCD) study and the Human Connectome Project (HCP) for adults, to understand the structural brain connectome and its relationship with cognition. Numerical results demonstrate huge advantages of GATE over competitors in terms of prediction accuracy, statistical inference and computing efficiency. We found that structural connectomes have a stronger association with a wide range of human cognitive traits than was apparent using previous approaches.