Modeling of individual neurophysiological brain connectivity

IntroductionComputational models are often used to assess how functional connectivity (FC) patterns emerge from neuronal population dynamics and anatomical connections in the brain. However, group averaged data is often used in this context and it remains unclear whether individual predictions of FC patterns using this approach can be made. Here, we assess the value of using individualized structural data for simulation of individual whole-brain FC. MethodsThe Jansen and Rit neural mass model was employed, where masses were coupled using individual structural connectivity (SC) obtained from diffusion weighted imaging. Simulated FC was correlated to individual magnetoencephalography-derived empirical FC. FC was estimated using both phase-based (phase lag index (PLI), phase locking value (PLV)) and amplitude-based (amplitude envelope correlation (AEC)) metrics to analyze the goodness-of-fit of different metrics for individual predictions. Prediction of individual FC was compared against the prediction of group averaged FC. We further tested whether SC of a different participant could equally well predict a participants FC pattern. ResultsThe AEC provided a significantly better match between individually simulated and empirical FC than phase-based metrics. Simulations with individual SC provided higher correlations between simulated and empirical FC compared to using the group-averaged SC. However, using SC from other participants resulted in similar correlations between simulated and empirical FC compared to using participants own SC. DiscussionThis work underlines the added value of FC simulations based on individual instead of group-averaged SC, and could aid in a better understanding of mechanisms underlying individual functional network trajectories in neurological disease. Impact statementIn this work, we investigated how well individual empirical functional connectivity can be simulated using the individuals structural connectivity matrix combined with neural mass modeling. Our research highlights the potential added value of using individual simulations of functional connectivity, and could aid in a better understanding of mechanisms underlying individual functional network trajectories in neurological disease. Moreover, individualized prediction of disease trajectories could enhance patient care and may provide better treatment options.