Neuro-Hotnet: A Graph Theoretic Approach for Brain FC Estimation

There has been increasing interest in the potential of multi-modal imaging to obtain more robust estimates of Functional Connectivity (FC) in high-dimensional settings. We develop novel algorithms adapting graphical methods incorporating diffusion tensor imaging (DTI) and statistically rigorous control to FC estimation with computational efficiency and scalability. Our proposed algorithm leverages a graphical random walk on DTI data to define a new measure of structural influence that highlights connected components of interest. We then test for minimum subnetwork size and find the subnetwork topology using permutation testing before the discovered components are tested for significance. Extensive simulations demonstrate that our method has comparable power to other currently used methods, with the advantage of greater speed, equal or more robustness, and simple implementation. To verify our approach, we analyze task-based fMRI data obtained from the Human Connectome Project database, which reveal novel insights into brain interactions during performance of a motor task. We expect that the transparency and flexibility of our approach will prove valuable as further understanding of the structure-function relationship informs the future of network estimation. Scalability will also only become more important as neurological data become more granular and grow in dimension.