Epilepsy region localization in signal domain — A smart getup

Epilepsy region localization is a key stage in pre-surgical evaluation of patients with medical refractory epilepsy (which is more burdensome than any other types of epilepsy). A common way of analyzing regions which are affected with epilepsy is using electroencephalogram by capturing electric potential from human scalp. In this article, a method is presented for the effective epilepsy region localization by analyzing electroencephalogram signals based on signal domain features. Features such as average signal energy and average log-energy entropy are extracted from signals and their differential signals in the signal domain. The experimental results show the ability of the proposed features to localize the epileptic regions by classifying the epileptic signals from non epileptic signals.