This study introduces an artificial neural network (ANN) trained to detect focal cortical dysplasias (FCDs), using the Morphometric Analysis Program (v2018; MAP18) for T1-weighted images. The ANN achieved high sensitivity and specificity across both training and independent validation datasets, demonstrating robust generalizability. This tool offers a clinically practical solution for improving presurgical evaluations in drug-resistant focal epilepsy.
Creator
University Hospital Bonn, Klinik Lengg AG
Information
Pediatrics or Adult
Adult
Speciality
Neurology
Modality
MRI
Training
113 patients with manually segmented FCDs and 362 healthy controls