Chronobiological Spatial Clusters of Cortical Regions in the Human Brain
Ravindra Arya,Gabrielle T Petito,Jeremy Housekeeper,Jason Buroker,Craig Scholle,Brian Ervin,Clayton Frink,Paul S Horn,Wei Liu,Marc Ruben,David F Smith,Jesse Skoch,Francesco T Mangano,Hansel M Greiner,Katherine D Holland
DOI: https://doi.org/10.1097/WNP.0000000000001119
2024-10-02
Abstract:Purpose: We demonstrate that different regions of the cerebral cortex have different diurnal rhythms of spontaneously occurring high-frequency oscillations (HFOs). Methods: High-frequency oscillations were assessed with standard-of-care stereotactic electroencephalography in patients with drug-resistant epilepsy. To ensure generalizability of our findings beyond patients with drug-resistant epilepsy, we excluded stereotactic electroencephalography electrode contacts lying within seizure-onset zones, epileptogenic lesions, having frequent epileptiform activity, and excessive artifact. For each patient, we evaluated twenty-four 5-minute stereotactic electroencephalography epochs, sampled hourly throughout the day, and obtained the HFO rate (number of HFOs/minute) in every stereotactic electroencephalography channel. We analyzed diurnal rhythms of the HFO rates with the cosinor model and clustered neuroanatomic parcels in a standard brain space based on similarity of their cosinor parameters. Finally, we compared overlap among resting-state networks, described in the neuroimaging literature, and chronobiological spatial clusters discovered by us. Results: We found five clusters that localized predominantly or exclusively to the left perisylvian, left perirolandic and left temporal, right perisylvian and right parietal, right frontal, and right insular-opercular cortices, respectively. These clusters were characterized by similarity of the HFO rates according to the time of the day. Also, these chronobiological spatial clusters preferentially overlapped with specific resting-state networks, particularly default mode network (clusters 1 and 3), frontoparietal network (cluster 1), visual network (cluster 1), and mesial temporal network (cluster 2). Conclusions: This is probably the first human study to report clusters of cortical regions with similar diurnal rhythms of electrographic activity. Overlap with resting-state networks attests to their functional significance and has implications for understanding cognitive functions and epilepsy-related mortality.