Dynamic characteristics of brain networks in patients with obsessive-compulsive disorder based on naturalistic paradigm.

Obsessive-compulsive disorder (OCD) is characterised by pervasive cognitive deficits, pointing to atypical organisation of large-scale functional brain networks. While such network abnormalities have been extensively documented in resting-state and task-based neuroimaging studies, far less is known about how the OCD brain dynamically reconfigures in response to naturalistic, context-rich stimuli. Here, we combined a naturalistic movie-watching paradigm with a Hidden Markov Model (HMM) to investigate the temporal evolution of brain network states under conditions that more closely approximate real-world cognitive demands. Eighty participants (35 with OCD and 45 healthy controls) underwent fMRI scanning during pre-movie rest, movie viewing, and post-movie rest. The HMM identified 14 distinct brain states, and group differences in their temporal properties were quantified using fractional occupancy, dwell time, and transition probability. While brain dynamics during movie watching were largely comparable across groups, marked abnormalities emerged in OCD during the post-watching resting state, including sustained activation of the default mode (DMN), salience (SAL), and executive control (EXEC) networks, alongside inflexible and dispersed state transitions. These altered dynamics were further associated with specific OCD symptom dimensions. Taken together, the findings suggest that context-dependent rigidity in DMN-SAL-EXEC dynamics may constitute a core neural mechanism of OCD and highlight potential intervention targets aimed at restoring flexible state transitions and adaptive network reconfiguration.