Early Prediction of Drug-Resistant Epilepsy using Clinical and EEG Features Based on Convolutional Neural Network

Epilepsy is a spontaneous and serious neurological disorder that presents with recurrent seizures and affects around 50 million people globally [1]. Unfortunately, despite recent advances in the development of antiseizure medications (ASMs), drug-resistant epilepsy (DRE) still affects 20% to 30% of patients with epilepsy (PWE) [1–3]. Patients with DRE bear significant economic, social, physical, and psychological burdens, but it takes a long time to identify DRE after repetitive ASMs trial, identifying patients at high risk of developing DRE early may select other treatment options, such as epilepsy surgery or neuromodulation or ketogenic diet earlier.

0