Shaylyn Kress

The paper compares diffusion and shortest path routing models for information transfer in the brain's structural network and finds the diffusion model better explains functional connectivity.

This paper develops a graph neural network to predict functional connectivity and centrality from structural connectivity, achieving high variance explanation in both mean and individual-level measures.

The paper investigates how brain structural connectivity, measured by DTI-derived centrality, relates to functional complexity in fMRI signals, finding that degree centrality best predicts the Hurst exponent.