EEG Brain Signal Processing for Epilepsy Detection

Abstract

Background: Millions of neurons make up the human brain, and they play an important role in controlling the body's response to internal and external motor and sensory stimuli. These neurons can function as contact conduits between the human body and the brain. Analyzing brain signals or photographs will help one better understand cognitive function. These states are linked to a particular signal frequency that aids in the comprehension of how a complex brain system works. Objective: Electroencephalography (EEG) is a useful method for locating brain waves associated with different countries on the scalp. Epilepsy is a condition where the brain or some part of it is overactive and sends too many signals. This results in seizures causing muscles to twitch or whole-body convulsions. Methods: In this paper, the author has designed a model to predict epilepsy using machine learning algorithms and deep learning models. For the machine learning algorithm, different features were extracted and a particle swarm optimization algorithm was used to select the best feature which was classified using wavelet transform.Vgg16, Vgg19, and Inception V3 models are used for the detection of epilepsy. Results: The inception V3 model results in 97.87% accuracy which is better than all other techniques. 5.1% accuracy improvement has been observed using a machine learning algorithm. The model is compared using existing work and it has been observed that the proposed model results better. Conclusion: The technique for modeling EEG signals and insight brain signals recorded during surgical procedures has been identified in detail. 0.7% and 0.13% accuracy improvement were achieved when the model is validated on Kaggle and CHB-MIT datasets respectively.