Morphometric analysis on T1-weighted MRI complements visual MRI review in focal cortical dysplasia

Abstract

ObjectiveFocal cortical dysplasia (FCD) is a common pathology in focal drug resistant epilepsy (DRE). Voxel based morphometric MRI analysis has been proposed as an adjunct to visual detection of FCD, which remains challenging given the subtle radiographic appearance of FCD. This study evaluates the diagnostic value of morphometric analysis program (MAP) in focal DRE with pathology-confirmed FCD. MethodsAutomated morphometric analysis program analysis generated z-score maps derived from T1 images, referenced to healthy adult or pediatric controls for each of 39 cases with pathology-confirmed FCD. MAP identified abnormal extension of gray matter into white matter (MAP-E) and blurring of the gray-white matter junction (MAP-J), independently of clinical data and other imaging modalities. MRI was visually reviewed by neuroradiologists as part of usual clinical care, and independently re-reviewed retrospectively by a neuroradiologist with >10-years’ experience in epilepsy MRI. Sensitivity and specificity were calculated for MRI, MAP, scalp-EEG, PET and SISCOM compared to resection area (RA). ResultsIn this cohort of 39 histologically proven FCD cases, the sensitivity and specificity of MAP-J [64% (95% CI 48%–77%) and 96% (95% CI 93%–0.98%)] and MAP-E [74% (95% CI 59%–86%) and 94% (95% CI 91%–97%)] were higher than qualitative MRI review, SISCOM, and FDG-PET. Initial MRI review detected FCD in 17, expert review identified 26. Among cases not detected by initial MRI review, MAP-J correctly identified FCD in 12 additional cases and MAP-E in 13 cases. Among cases not detected by expert MRI review, MAP-J correctly identified 6 and MAP-E 8 cases. Excellent surgical outcome was achieved in 76% of patients. SignificanceMAP showed favorable sensitivity compared to visual inspection and other non-invasive imaging modalities. MAP complements non-invasive imaging evaluation for detection of FCD in focal DRE patients.