Focal cortical dysplasia (FCD) is a common pathology underlying medically intractable focal epilepsies. Conventional MRI can be limited in characterizing subtle FCD, due to the lack of quantitative measurements for tissue properties. Here, we proposed a multiparametric machine learning (ML) approach based on high-resolution 3D MR fingerprinting (MRF) to characterize FCD lesions. The ML model showed robust accuracy of 96%, 89%, and 79% to separate FCD from normal cortex, FCD type I from type II, and FCD type IIa from type IIb, respectively. Our findings suggest the usefulness of the multiparametric MRF ML approach to improve noninvasive epilepsy presurgical evaluation.
This abstract and the presentation materials are available to members only; a login is required.