Data-driven brain network models predict individual variability in behavior

TL;DR

This study uses personalized brain network models based on individual structural connectivity to predict how brain differences influence task performance and response to stimulation, advancing personalized neuroscience.

Contribution

It introduces a data-driven computational approach combining anatomical connectivity and nonlinear brain dynamics to predict individual behavioral variability.

Findings

Task performance correlates with local or global brain activity measures.

Structural connectivity differences influence brain dynamics and behavior.

Model predicts individual responses to targeted brain stimulation.

Abstract

The relationship between brain structure and function has been probed using a variety of approaches, but how the underlying structural connectivity of the human brain drives behavior is far from understood. To investigate the effect of anatomical brain organization on human task performance, we use a data-driven computational modeling approach and explore the functional effects of naturally occurring structural differences in brain networks. We construct personalized brain network models by combining anatomical connectivity estimated from diffusion spectrum imaging of individual subjects with a nonlinear model of brain dynamics. By performing computational experiments in which we measure the excitability of the global brain network and spread of synchronization following a targeted computational stimulation, we quantify how individual variation in the underlying connectivity impacts both local and global brain dynamics. We further relate the computational results to individual variability in the subjects' performance of three language-demanding tasks both before and after transcranial magnetic stimulation to the left-inferior frontal gyrus. Our results show that task performance correlates with either local or global measures of functional activity, depending on the complexity of the task. By emphasizing differences in the underlying structural connectivity, our model serves as a powerful tool to predict individual differences in task performances, to dissociate the effect of targeted stimulation in tasks that differ in cognitive complexity, and to pave the way for the development of personalized therapeutics.