Dynamical models reveal anatomically reliable attractor landscapes embedded in functional brain networks

Poster Session C - Sunday, April 14, 2024, 5:00 – 7:00 pm EDT, Sheraton Hall ABC

Ruiqi Chen¹ (chen.ruiqi@wustl.edu), Matthew Singh¹, Todd Braver¹, ShiNung Ching¹; ¹Washington University in St. Louis

Analysis of resting state brain networks (RSNs) has generated many insights into cognition. However, the mechanistic underpinnings of resting state dynamics and RSNs are still not well-understood. It remains debated whether resting state activity is best characterized as noise-driven fluctuations around a single stable state, or instead, in terms of a nonlinear dynamical system with nontrivial attractors embedded in the RSNs. Here, we provide evidence for the latter by constructing whole-brain dynamical systems models from individual resting-state fMRI recordings in the Human Connectome Project using the Mesoscale Individualized NeuroDynamic (MINDy) platform. The MINDy models consist of hundreds of neural masses representing brain parcels, connected by fully trainable and individualized weights. The models revealed a diversity of nontrivial attractor landscapes including multiple equilibria and limit cycles. However, when projected into anatomical space, these attractors mapped onto a limited set of canonical RSNs, including default mode network and cognitive control network, which were reliable at the individual level. Further, by creating convex combinations of models, several bifurcations were induced that recapitulated the full spectrum of dynamics found via fitting. These findings suggest that the resting brain traverses a quite diverse set of dynamics, which generates several distinct but anatomically overlapping attractor landscapes. Thus, the spatiotemporal structure of the resting brain may be better captured through neural dynamical modeling and analysis. In ongoing research, we extend MINDy to cognitive task states to reveal task-triggered reconfiguration of whole-brain dynamics, providing a new analysis framework for cognitive neuroscience that connects rest and tasks.

Topic Area: METHODS: Neuroimaging