Abstracts

Rationale: Electroencephalographic source imaging (ESI) is becoming a useful tool for pre-surgical epilepsy workup. Various studies have examined the usage of different resolution head models. However, accurate ESI hasn't been realized broadly in clinical evaluations due to insufficient surface sampling and inability to generate computationally-demanding high-quality individual head models, such as finite difference models (FDM). This study aims to discuss the feasibility and accuracy of dense-array ESI (dESI) based on high-quality FDM (dESI-FDM). Methods: Candidates for resective epilepsy surgeries were selected for this study, including either one-stage or staged surgeries. dESI-FDM was realized with high-resolution surface potential data acquired by 256-channel dense-array electroencephalography (EEG) (Electrical Geodesics, Inc., USA) and high-quality FDM constructed from structural magnetic resonance imaging (MRI) (workflow of dESI-FDM, see Figure 1). Numerically, FDM specifying conductivity of multiple head contents (gray matter, white matter, cerebral spinal fluid, skull, eyeballs, air, and flesh) would be built. EEG sensor positions (average positions were applied in this study) were then registered to each individual's head model before computation of the lead field, with the following conductivity values: gray matter =0.25 S/m, white matter =0 .35 S/m, CSF = 1.79 S/m, skull = 0.01 S/m, eyeball = 1.55 S/m, and flesh = 0.33 S/m. Typically 2447 oriented dipoles (perpendicular to cortical surface) are distributed in each hemisphere. Standardized low-resolution brain electromagnetic tomography (sLORETA) was selected as the inverse method. Validity of dESI-FDM results from interictal and/or ictal epileptic activities were testified by multiple standards including the "gold standard" intracranial EEG (icEEG). Results: Fifteen patients who achieved excellent outcome (Engel grade I) after resective surgeries were included in this study. dESI-FDM provided sufficient accuracy in 93.3% cases according to multiple validating standards of "icEEG" (see Figure 2, an illustrative case), "resected area" or "proximity to resected lesions" for staged surgeries, expanded resections or lesionectomies (limited resections), respectively. dESI-FDM results also contributed to the planning of all the intracranial electrodes implantation surgeries. Accurate dESI-FDM was successfully realized in five difficult cases despite the patients' unsymmetrically misshapen head shapes or large cerebral lesions which requires experienced manual editing of the head model. Conclusions: With high-quality FDM, sufficient surface potential sampling and proper inverse methods, accurate dESI is proven to be feasible and contributable to pre-surgical workup of epilepsy surgeries, even in difficult cases with abnormal head shapes. More validation studies involving comparison between "gold standard" icEEG and dESI-FDM results (in at least the level described in this study) should be performed in future to further consolidate dESI-FDM's accuracy and make it contribute more to pre-surgical workup of epilepsy. Funding: None