Beyond structural MRI: combined diffusion metric improve delineation of epileptogenic tissue in focal cortical dysplasia

Abstract

Abstract Accurate delineation of the epileptogenic zone in focal cortical dysplasia (FCD) remains challenging, as structural MRI often underestimates the extent of dysplastic tissue. Diffusion MRI can probe microstructural alterations, yet individual diffusion models provide limited sensitivity and clinical interpretability. We prospectively studied 19 patients with histologically confirmed FCD who underwent presurgical evaluation, surgical resection, and postsurgical MRI. Fourteen diffusion metrics derived from complementary diffusion models were integrated using logistic regression to generate a single combined diffusion metric (CDM). Spatial correspondence among structural MRI, resected tissue, perilesional cortex, and contralateral homologues was assessed using Dice similarity and ROC analyses. Histological cellularity and multimodal imaging data were used for biological validation. Individual diffusion metrics identified microstructural abnormalities within MRI-visible lesions, including reduced anisotropy, lower fibre density, and increased isotropic diffusion. Integration into the CDM improved discrimination between lesional and healthy tissue ( p  < 0.03), detected abnormalities extending beyond radiological borders, and showed spatial concordance with FDG-PET hypometabolism and epileptiform activity in 90% of patients. CDM-guided delineation increased agreement with resected regions and reduced unresected perilesional volume. CDM provides a biologically grounded and clinically interpretable representation of epileptogenic tissue in FCD, refining presurgical mapping beyond conventional MRI and potentially improving surgical targeting.