Generalized Tonic-Clonic Seizures Detection Using Deep Learning Techniques

Epilepsy treatment can be significantly enhanced through automated seizure detection from electroencephalography. This study focuses on the detection of generalized tonic-clonic seizures, a critical seizure type associated with risks such as sudden unexpected death in Epilepsy (SUDEP) and postictal pulmonary edema (PPE), by leveraging advanced deep learning models, including Self-Supervised Graph Neural Networks, Long Short-Term Memory networks (LSTM) and CNN. This research aims to improve the precision and reliability of detecting tonic-clonic seizures, as well as applying techniques such as data augmentation and specialized loss functions. These novel approaches demonstrate promising results in enhancing the detection capabilities of EEG-based seizure detection systems.