Brain Network Identification in Asynchronous Task fMRI Data Using Robust and Scalable Tensor Decomposition

摘要

The goal of this work is to robustly identify common brain networks and their corresponding temporal dynamicsacross subjects in asynchronous task functional MRI (tfMRI) signals. We approached this problem using a robust andscalable tensor decomposition method combined with the BrainSync algorithm. We first used BrainSync algorithm totemporally align asynchronous tfMRI data, allowing us to study common brain networks across subjects. We mapped thesynchronized tfMRI data into a 3D tensor (vertices × time × session) and performed a greedy canonical polyadic (CP)decomposition, reducing the rank to 20 in order to improve the signal-to-noise ratio (SNR). We incorporated the Nesterovacceleratedadaptive moment estimation into our previously developed scalable and robust sequential CP decomposition(SRSCPD) framework and applied this improved version of SRSCPD to the rank-reduced tensor to identify dynamic brainnetworks. We successfully identified 9 brain networks with their corresponding temporal dynamics from 40 subjects usingHuman Connectome Project tfMRI data without using any prior information with regard to the task designs. Three of theseshow the subjects’ responses to cues at the beginning of each task block (fronto-parietal attentional control network, visualnetwork and executive control network); one corresponds to the default mode network that exhibits deactivation duringthe tasks; four show motors networks (left hand, right hand, tongue, and both feet) where the temporal dynamics arestrongly correlated to the task designs, and the remaining component reflects physiological noise (respiration).