The cyclical nature of epilepsy has been a growing area of research in the last several years. Implanted devices providing chronic recordings of brain activity have enabled the longitudinal study of interictal spiking and seizures over months to years. Leveraging knowledge of cycles of epileptic brain activity may enable seizure risk forecasting.

Epilepsy is increasingly understood as a cyclical disorder characterized by rhythms at multiple timescales, including ultradian, circadian, and multidien cycles. In this study, we utilize advanced mathematical tools to characterize the inter- and intra-subject variability in these cycles.

Using retrospective cohort study methods, we investigate interictal epileptiform activity (IEA) in patients implanted with the NeuroPace RNS^® System. We compare the utility of several mathematical tools in decomposing multidien rhythms and quantifying their variability within and across patients. 

We found that patients demonstrating multidien rhythms of IEA can be grouped into several chronotypes based on their dominant multidien rhythms. Some patients have highly robust multidien rhythms, while others are more variable. Characterizing patient-specific tendencies will be essential for personalized seizure risk assessment.

Multidien rhythms of IEA are patient-specific and range from highly stable to more variable.  Our findings motivate future studies of cycles of epileptic brain activity and their potential to inform clinical management of patients with epilepsy.