Litcius/Paper detail

Pioneering Seizure Prediction: Exploring ML and DL Approaches with IEEG Data

Veera Venkata Raghunath, Dedeepya Sai Gondi, Jubin Thomas, Hemanth Volikatla

202415 citationsDOI

Abstract

Epilepsy is a neurological condition with the brain which makes people has sudden—uncontrolled shaking called seizures. Some people have epilepsy that even medicine cannot help. However, scientists are now interested in using Machine Learning (ML) and Deep Learning (DL) to predict when seizures will happen. This could help stop seizures before they start by using devices called neuro-stimulators. But the methods used before need lots of data from brain tests such as— Electroencephalograms (EEG). Getting this data is hard—and it might not show what really happens in real life. Therefore, this study looks at a lot of different ways to use ML and DL to predict seizures. We use data from inside the brain called Intracranial Electroencephalography (IEEG). We tries out many different models—including new ones that use graphs, to see which one works the best for predicting seizures. Unlike other studies—this one uses only a little bit of data to test the models. We show that ML could be really useful for predicting seizures. The results of these tests show how well each model works and give advice on how to deal with this important health problem.

Topics & Concepts

Computer scienceArtificial intelligenceEEG and Brain-Computer InterfacesMachine Learning in HealthcareEpilepsy research and treatment