Activation network improves spatiotemporal modelling of human brain communication processes

• DOI: 10.1016/j.neuroimage.2023.120472
• Publication Date: 2023-11-23T23:15:18Z
• Authors: Xucheng Liu; Ze Wang; Shun Liu; Lianggeng Gong; Pedro A. Valdes Sosa; Benjamin Becker; Tzyy-Ping Jung; Xi-jian Dai; Feng Wan
• Keywords: Dynamic functional connectivity; Dynamic network analysis

Dynamic functional networks (DFN) have considerably advanced modelling of the brain communication processes. The prevailing implementation capitalizes on system and network-level correlations between time series. However, this approach does not account for continuous impact non-dynamic dependencies within statistical correlation, resulting in relatively stable connectivity patterns DFN over with limited sensitivity dynamic regions. Here, we propose an activation network framework based activity (AFC) to extract new types during process. AFC captures potential time-specific fluctuations associated processes by eliminating dependency correlation. In a simulation study, positive correlation (r=0.966,p<0.001) extracted simulated "ground truth" validates method's detection capability. Applying autism spectrum disorders (ASD) COVID-19 datasets, proposed extracts richer topological reorganization information, which is largely invisible DFN. Detailed, exhibits significant inter-regional connections function-specific subnetworks reconfigures more efficiently temporal dimension. Furthermore, fails distinguish patients healthy controls. method reveals decrease (p<0.05) information processing abilities patients. Finally, combining two successfully classifies ASD (83.636 % ± 11.969 %,mean±std) (67.333 5.398 %). These findings suggest could be analytic elucidating neural mechanism dynamics.