DeepMultiConnectome: Deep Multi-Task Prediction of Structural Connectomes Directly from Diffusion MRI Tractography

TL;DR

DeepMultiConnectome is a deep learning model that rapidly predicts brain structural connectomes directly from tractography data, eliminating the need for time-consuming gray matter parcellation and supporting multiple schemes.

Contribution

It introduces a multi-task neural network that directly predicts connectomes from tractography, significantly speeding up the process and maintaining high accuracy across different parcellation schemes.

Findings

Predicts connectomes in approximately 40 seconds from large tractograms.

Achieves high correlation with traditional connectomes ($r > 0.98$).

Maintains reproducibility and predictive performance for age and cognition.

Abstract

Diffusion MRI (dMRI) tractography enables in vivo mapping of brain structural connections, but traditional connectome generation is time-consuming and requires gray matter parcellation, posing challenges for large-scale studies. We introduce DeepMultiConnectome, a deep-learning model that predicts structural connectomes directly from tractography, bypassing the need for gray matter parcellation while supporting multiple parcellation schemes. Using a point-cloud-based neural network with multi-task learning, the model classifies streamlines according to their connected regions across two parcellation schemes, sharing a learned representation. We train and validate DeepMultiConnectome on tractography from the Human Connectome Project Young Adult dataset ($n = 1000$), labeled with an 84 and 164 region gray matter parcellation scheme. DeepMultiConnectome predicts multiple structural connectomes from a whole-brain tractogram containing 3 million streamlines in approximately 40 seconds. DeepMultiConnectome is evaluated by comparing predicted connectomes with traditional connectomes generated using the conventional method of labeling streamlines using a gray matter parcellation. The predicted connectomes are highly correlated with traditionally generated connectomes ($r = 0.992$ for an 84-region scheme; $r = 0.986$ for a 164-region scheme) and largely preserve network properties. A test-retest analysis of DeepMultiConnectome demonstrates reproducibility comparable to traditionally generated connectomes. The predicted connectomes perform similarly to traditionally generated connectomes in predicting age and cognitive function. Overall, DeepMultiConnectome provides a scalable, fast model for generating subject-specific connectomes across multiple parcellation schemes.