Vision Transformer for Epilepsy Seizure Prediction and Detection: A Narrative Review

Epilepsy could be a constant neurological disease stamped by recurrent seizures. World Health Organization (WHO) stated that around 50 million people in the world are affected by epileptic seizure. Affective seizure prediction and discovery are vital for overseeing epilepsy and improving the well-being of those affected. While different techniques have been recommended for this disease, there's a developing demand for tackling the capabilities of vision transformers. Vision transformers, which is a variation of artificial neural networks, have displayed promising results in areas like image classification and object detection in computer vision. As of now, Research is conducted in exploring their potential for predicting and detecting epilepsy seizures. Typically, due to their capacity to capture spatial and temporal designs inside electroencephalogram data. This review points to carefully examine the recent research on utilizing vision transformers for epilepsy seizure prediction and detection. The measure for selecting relevant studies is whether the paper is related to Epilepsy prediction or detection utilizing EEG information or not. Through a comprehensive assessment of relevant works, we dig into the strategies, performance metrics, preferences, and limitations in existing vision transformer models in this setting. We hope that by tackling the capabilities of vision transformers, it is possible to develop more exact and reliable models for estimating and recognizing epileptic seizures. We hope our work will further help researchers in their studies to find better solutions.