Title: A sparse blind source separation method for probing human whole-brain connectomes
Authors: Ying Guo - Emory University (United States) [presenting]
Yikai Wang - Emory University (United States)
Abstract: In neuroscience research, imaging-based network connectivity measures have become the key for understanding brain connectomes, potentially serving as individual neural fingerprints. There are major challenges in analyzing connectivity matrices including the high dimensionality of brain networks, unknown latent sources underlying the observed connectivity, and the large number of brain connections leading to spurious findings. We propose a novel blind source separation method with low-rank structure and uniform sparsity (LOCUS) as a fully data-driven decomposition method for network measures. Compared with existing methods that vectorizes connectivity matrices ignoring brain network topology, LOCUS achieves more efficient and accurate source separation for connectivity matrices using the low-rank structure and a novel angle-based uniform sparsity regularization. We propose an efficient iterative Node-Rotation algorithm to solve the non-convex optimization problem for learning LOCUS. We illustrate LOCUS through extensive simulation studies and application to a resting-state fMRI data.