Effort and yield of implementing electrical source imaging in presurgical epilepsy evaluation

Abstract-Text (inklusive Referenzen und Bildunterschriften)

Introduction: Electric source imaging (ESI) is a computational analysis method integrating scalp electroencephalography (EEG) and MRI, that assists in localizing EEG activity within a 3D model of the brain. Although ESI has been established as an accurate and clinically helpful diagnostic tool in presurgical evaluation of patients with drug-resistant epilepsy, it is underutilized in European epilepsy centers. Possible barriers to integration of ESI into clinical practice include perceived high costs and the need for extensive specialized training, however, these have not yet been systematically examined.

Goals: To (1) investigate the cost in working hours of ESI integration into an epilepsy surgery program and (2) examine concordance of results from freely available and commercial ESI software programs.

Methods: This prospective study included all drug-resistant epilepsy patients referred for noninvasive presurgical evaluation to a epilepsy surgery center from July 2021 to July 2022. Interictal ESI was analyzed from 33- to 45-electrode EEG recorded during long-term video EEG monitoring and modelled onto individual head models using distributed inverse solutions. Working hours required for ESI set up and processing of both the treating epileptologist and IT personnel were recorded, as well as concordance of results between two freely available academic (Brainstorm, Cartool) and one commercial, CE-certified software (Epilog) programs. Additionally, a survey of all tertiary epilepsy centers in Germany was conducted to assess the current state of ESI integration in presurgical practice.

Results: Of 40 patients included during the 1-year study period, 22 (53%) had interictal activity sufficient for ESI analysis. A total of 35.6 and 33.0 hours were needed for initial IT ESI set up and physician training, respectively. Following initial physician training, an average 3.3 ± 2.1 physician hours were required per analyzed case. Working times became significantly shorter as study period progressed, with the last two-thirds of patients requiring an average 2.3 ± 0.5 hours per patient. Comparing time needed for different software programs, the commercial software had shorter processing times with 0.4 ± 0.3 hours per patient compared to the two freeware packages requiring 1.0 ± 0.8 and 0.9 ± 0.7 hours per patient, p = .003. Complete concordance of ESI localization on the sublobe level between all three software programs was 22.7%, reflecting low agreement. Finally, a survey of all 19 German epilepsy centers with epilepsy surgery programs revealed that 53% of centers use ESI in presurgical practice; however, only 26% regularly use ESI results for clinical decision making.

Summary: This study provides information regarding expected effort and costs for integration of ESI into a surgical epilepsy center. Low concordance between different ESI programs suggests standardized guidelines for ESI implementation are needed.