Clinical utility of semi-automated EEG electric source imaging of interictal discharges in presurgical evaluation and surgical treatment decision making

Jelena Hyppönen^1,2*Henri Eronen^2,3Anni Saarela^2,4Päivi Koskenkorva⁵Esa Mervaala^1,2Reetta Kälviäinen^2,3Leena Jutila³

• ¹Kuopio Epilepsy Center, Full Member of ERN EpiCARE, Department of Clinical Neurophysiology, Kuopio University Hospital, Kuopio, Finland
• ²Institute of Clinical Medicine, School of Medicine, Faculty of Health Sciences,, University of Eastern Finland, Kuopio, Northern Savonia, Finland
• ³Kuopio Epilepsy Center, Full Member of ERN EpiCARE, Department of Neurology, Neurocenter, Kuopio University Hospital, Kuopio, Northern Savonia, Finland
• ⁴Department of Child Neurology, Kuopio Epilepsy Center, Neurocenter, Kuopio University Hospital, Full Member of ERN EpiCARE, Kuopio, Finland, Kuopio, Finland
• ⁵Full Member of ERN EpiCARE, Deparment of Clinical Radiology, Diagnostic Imaging Center, Kuopio University Hospital, Kuopio, Northern Savonia, Finland

We aimed to evaluate the clinical utility of the semi-automated EEG source imaging (ESI) analysis results during presurgical workup in epilepsy surgery candidates.: Forty consecutive presurgical patients with focal epilepsy, evaluated during 2021-2022, with a mean age of 35.0 years (range, 2-59), were included. The long-term video-EEG recordings were analyzed using Epilog PreOp (Clouds of Care, Ghent, Belgium). Automatically detected spike clusters were evaluated, and source localization at the half-rising phase of the averaged spike peak was used for ESI localization at the lobar level. The evaluation by the multidisciplinary epilepsy surgery team (MDT), including clinical and multimodal presurgical evaluation data, was used to assess the clinical utility of ESI results in the decision-making process.Results: Twenty of 40 patients were classified as non-lesional based on magnetic resonance imaging (MRI), and 22 of 36 had focal PET hypometabolism. In automated ESI analyses, no epileptic spikes were detected in four patients (10%). The largest spike clusters were localized to the temporal (58%), frontal (23%), and occipital lobes (3%) or the insula (8%). ESI results had an impact on the suggested treatment choice in 45% of cases and did not significantly affect clinical decision-making in 35% of cases. In patients who underwent resective surgery or had MRI lesions in the temporal lobe, ESI findings were more likely to be consistent with other findings and to complement clinical decisions (Pearson chi-square p = 0.016 and 0.027).This study demonstrates that commercial automated spike detection analyses and source localization, when reviewed by an experienced MDT, can aid clinical decision-making in about 40% of patients, especially for patients in whom resective epilepsy surgery is a feasible option.