An Early Detection Technique for the detection of Epilepsy using EEG signals

Abstract

The advent of the EEG signals and their applications in medical filed have opened many new avenues for the mankind to excel in. Today Electroencephalogram or EEG is one of the frontline tool for diagnosis of brain related issues. EEG plays the pivotal role in making the whole neurological disorder mapping more easy and time efficient. Through the state of the art diagnosis techniques thanks to advancement in pattern recognition and artificial intelligence systems are being modeled which can detect any kind of neurological disorders and can make appropriate changes into them as well. EEG signals demonstrate the electrical movement of brain and consist of useful data for the different states of brain to get and study the detailed information about the brain. The identification of various categories of EEG signals is normally performed by the experts from the field of visual inspection. To complete that process manually it may cause human based errors and cause a lot of time and also considered as time taking process and also not much reliable. Visually inspecting the EEG signals for the purpose of detecting the epileptic seizures varies according to the expertise of human experts. That’s why, an automatically detection of epileptic seizures is essential in the real environment. Our proposed methodology is evaluated with different parameters and testing accuracy of 94 % is reported for a publicly available dataset.to maintain the state of the art technique is applied on one other dataset and in that Total training accuracy of 98.74% is obtained through that and testing accuracy which is obtained is 95.22%