Abstracts

Rationale: The evaluation, diagnosis and surgical treatment of patients with intractable cingulate epilepsy is difficult, owing to its diverse semiology, complex organization of implicated networks and surgical challenges in accessing this region of the brain. Stereoelectroencephalography (sEEG) has improved the ability to identify and localize the epileptogenic zone (EZ) in cingulate epilepsy, although high-dimensional sEEG datasets are still largely qualitatively reviewed by clinicians. The success of surgical intervention in cingulate epilepsy relies on more accurate EZ characterization using quantitative metrics highlighting the inherent complexity of spatiotemporal dynamical features in this data. Phase-amplitude coupling (PAC) has emerged as a promising signal biomarker and estimate of dynamical network synchronization between recorded sites. We hypothesize here that PAC may be a useful tool for sEEG analysis by identifying networks implicated in cingulate epilepsy.


Methods: We identified four seizure-free patients with complex refractory cingulate epilepsy who were treated at the Cleveland Clinic Epilepsy Center. All patients underwent sEEG for invasive monitoring (Table 1), followed by laser ablation and/or surgical resection, achieving a minimum of 1 year of seizure-freedom. PAC analysis was applied to each sEEG electrode contact for pre-ictal and ictal epochs across all seizures using 10-second window segments, with 2-Hz bandwidth spacing for low-frequency (1-13 Hz) and 5-Hz spacing for high-frequency signals (14-250 Hz). Average PAC values in resected (more highly epileptogenic regions) and non-resected contact sites were compared using statistical analysis (Mann-Whitney U-test).


Results: Our study reveals statistically significant variances (p< 0.05) in PAC values between resected (more highly epileptogenic regions) and non-resected sEEG electrode contacts across the pre-ictal and ictal states. PAC also facilitated EZ lateralization in this diagnostically challenging patient cohort with mesial onset cingulate epilepsy, further supporting PAC as a biomarker to aid in EZ network identification.


Conclusions: Despite the complexity of cingulate epilepsy and challenges with sEEG interpretation in this cohort of patients, this study supports the application of PAC analysis as a potential feature helping to identify electrode contacts implicated in the EZ. Our findings support further research using the PAC method across sEEG recordings to characterize personalized, patient-specific epileptiform activity in brain networks including the cingulate region, and offer insights that may help with surgical decision-making for epilepsy surgery.


Funding: None.