Sophia Robert¹, Christina Patterson², Marlene Behrmann^2,1; ¹Carnegie Mellon University, ²University of Pittsburgh

After unilateral cortical resection in childhood, the preserved hemisphere must support perception and cognition with markedly altered structural connectivity. Time-varying functional connectivity can reveal how the preserved hemisphere coordinates activity moment-to-moment and whether the properties of transient high-amplitude network co-fluctuations are altered. We used movie-watching fMRI and an edge-centric co-fluctuation framework to test whether whole-hemisphere dynamics and properties of high-amplitude (HA) events differ from controls after surgery. Eight patients with unilateral cortical resections for drug-resistant epilepsy during childhood (15–20 years; 3 male; 5 with a preserved left and 3 with a preserved right hemisphere) and seven typically developing controls (9–20 years; 2 male) watched a 10.5-minute film segment twice on a 3T MRI scanner. Cortical time series from the preserved hemisphere were parcellated (Schaefer-400) and converted to edge time series. The root-sum-of-squares (RSS) across edges indexed global co-fluctuation amplitude at each frame. HA events were defined as contiguous low-motion frames where RSS exceeded a permutation-derived threshold. We quantified intersubject similarity of global and network-restricted RSS trajectories and compared HA-event rate and amplitude between groups using mixed-effects models, complemented by single-case tests with control-as-patient validation. Global and network-level RSS intersubject correlations showed substantial heterogeneity across individuals and no large group shifts in this sample. In contrast, group models indicated reduced HA-event rate (but not amplitude), and case-series analyses showed this effect was driven by a subset of patients with an intact left hemisphere whose event rates were markedly below the hemisphere-matched control distribution, classified as highly robust abnormalities. Other patients fell within or close to the control range, and almost all showed normative event amplitudes, even when rates were reduced. These results indicate lower rates of high-amplitude co-fluctuation events in the contralesional hemisphere after pediatric cortical resection and underscore the value of case-level analyses for characterizing post-surgical large-scale network dynamics.

Acknowledgements: This research was conducted at the CMU-Pitt BRIDGE Center (RRID: SCR_023356) and supported by the NIH, National Eye Institute (NEI, R01EY027018), the NSF GRFP (DGE2140739), an NEI P30 CORE Award (EY08098), The Research to Prevent Blindness Inc, NY, and the Eye & Ear Foundation of Pittsburgh.