Abstracts

Rationale: The purpose of this study is to evaluate if diffusion tensor imaging (DTI) can predict epileptogenic tubers by measuring apparent diffusion coefficient (ADC), fractional anisotropy (FA), and radial diffusivity (RD) of tubers and perituber tissues, and to investigate if myelination interferes with the detection of the epileptogenic tissue changes. Methods: Twenty-two patients (0.4-19.6 year old, mean age 5.3; 12 female, 10 male) who underwent tuber resection for treatment of epilepsy between 2004 and 2011 were retrospectively selected from UCLA TSC cohort of our institute. All patients had preoperative DTI. Total of 545 tubers were divided into two groups; epileptogenic and non-epileptogenic. Epileptogenicity was defined by preoperative evaluation of electroencephalography (EEG), ¹⁸F-2-fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET), Magnetoencephalography, and electrocorticography. The data analysis was performed using Analysis of Functional NeuroImages (AFNI). Two observers manually outlined all tubers (ROI^tuber) by consensus on ADC map with a reference of T2WI or FALIR. 4 mm thick ring-shaped ROIs of perituberal tissue (ROI^perituber) and ROIs including tuber and perituberal tissue (ROI^{tuber+perituber}) were also generated. Maximum, minimum, mean, and median values of ADC, FA, and RD were calculated. Mann-Whitney U test and unpaired t test were used to compare these DTI parameters between epileptogenic and non-epileptogenic tubers. To investigate the influence of white matter myelination, a group of patients younger than 3 was separately analyzed and results were compared with all patients. Results: Maximum ADC and RD were significantly higher in epileptogenic tubers. FA showed no significant difference between groups. The changes were more significant in patients younger than 3. Furthermore, ROI^tuber and ROI^{tuber+perituber} showed higher mean ADC and lower minimum FA in epileptogenic tubers. Maximum ADC value of ROI^{tuber+perituber} showed most significant difference in all patients and patients younger than 3 (Fig. 1a, 1b). ROC curves showed higher value of area under curve in patients younger than 3 (Fig. 2a, 2b). Conclusions: ADC and RD of tuber and tuber + perituberal tissue were higher in epileptogenic group. DTI can be a promising additional tool to predict epileptogenic tissue, especially in patients younger than 3 years old.