Project Summary: Overview: Neurosurgical resection, ablation and stimulation can be curative for patients with medically refractory epilepsy (MRE), however, successful treatment depends on precise localization of the seizure onset zone (SOZ). Stereo-electroencephalography (EEG) is the gold standard for determining the SOZ and involves implanting many intra-cranial depth electrodes to detect epileptiform activity. Unfortunately, there are often no visible lesions on brain MRI and only a crude spatial map from scalp-EEG to guide the stereo-EEG implantation, resulting in electrode positions that fail to definitively determine the SOZ. To meet the unmet clinical need for consistent and accurate stereo-EEG guidance we will develop innovative diffusion MRI (dMRI) encoding paradigms that provide specific sensitivity to epileptic microstructural pathology and validate with 3D histology and high-density scalp- and stereo-EEG. Relevance: There are ~400,000 patients with MRE and non-lesional MRI in the U.S. Given the lack of guidance and sparse sampling of stereo-EEG, identifying the seizure onset zone with confidence is extremely challenging for in these non-lesional patients. If we are successful in the proposed work, we will enable MRI identification of a probable lesion location for the majority of patients with MRE. In this way , we will increase their likelihood of a definitive stereo-EEG identification of the SOZ and enable a potentially curative neurosurgical treatment option that might not have been possible otherwise. Approach: Our approach is to: 1) Develop and validate diffusion MRI cortical fiber mapping for detection of the disrupted cortical architecture within focal cortical dysplasias. 2) Test whether diffusion MRI cortical fiber mapping can detect focal cortical dysplasias and 3) Test whether diffusion tractography and functional connectivity MRI can predict seizure propagation. Specifically, we will use these MRI connectivity to predict probable cortical nodes of the seizure propagation network, as well as, the latency and spatial propagation of epileptiform activity between stereo-EEG electrodes. Summary: The proposed neuroimaging methods will improve localization of the SOZ and propagation network enabling more patients with MRE to be treated more effectively with neurosurgery.