신경 신호분석을 통한 난치성 뇌전증환자의 발작 기시부위 국지화 연구

Abstract

Traditionally, identification of epileptogenic zones primarily relied on visual inspection of intracranial electroencephalography (iEEG) recordings by experienced epileptologists; however, removal of epileptogenic zones identified using iEEG does not always guarantee favorable surgical outcomes. To confirm the visual inspection results and assist in making decisions about surgical resection areas, computational iEEG analysis methods have recently been used for the localization of epileptogenic zones. In this study, I have proposed a new approach for the localization of epileptogenic zones in Lennox–Gastaut syndrome (LGS) and have investigated whether the proposed approach could be a promising tool to confirm surgical resection areas and to predict the seizure outcome before the surgery. The proposed approach simultaneously used results of two iEEG analysis methods, betweenness centrality (BC) and directed transfer function (DTF) analyses, to enhance overall localization accuracy. This new combinatory method was applied to patients who became seizure-free after resective epilepsy surgery as well as those who had unsuccessful surgeries. The estimated epileptogenic zones more strongly coincided with surgical resection areas in patients with favorable surgical outcomes compared to those with unsuccessful surgical outcomes. The qualitative analysis showed that the combinatory use of two iEEG analysis methods might result in a more accurate estimate of epileptogenic zones in LGS than the use of a single analysis method. An appropriate combination of multiple iEEG analysis methods could not only enhance the overall accuracy of localizing epileptogenic zones in LGS but also potentially be used to predict surgical outcomes of patients with LGS before resective surgery.