Epilepsy is a chronic brain disorder and epileptic patients encounter recurrent seizures caused by abnormally synchronous electrical activity in parts of the brain. Over 50 million people spread across the world have epilepsy amongst whom approximately 30% suffer from refractory epilepsy which cannot be controlled by existing treatment protocols. For all epileptic sufferers, the thought that their next seizure could come at any time is agonizing and traumatic. However, if seizures could be predicted reliably, associated dangers and inconveniences will be greatly mitigated. Although the epileptic seizure prediction challenge has been tackled headlong by researchers through different modelling methods the problem of prediction has not yet been satisfactorily solved. In this paper, a systematic literature review of prominent epileptic seizure prediction attempts was carried out. We focus majorly on the two predominant classes of modelling attempts used: physiological mechanism and data based. The review underscores the richness and utility of the diverse modeling strategies as well as the gainful contribution of researchers in the field of epilepsy. It shows that meaningful progress has been made towards discovering the exact mechanism of seizure generation and realization of reliable and consistent seizure prediction algorithm
A Review of Epileptic Seizure Prediction: Physiological Mechanism and Data Based Attempts
