Machine Learning-Based Epileptic Seizure Detection Using XGboost Algorithm

It is estimated that around 50 million people around the world have epilepsy, with approximately 80% of the people living in low- and middle-income countries. Many people with epilepsy may not be formally diagnosed or may not report their condition due to social stigma or other barriers to healthcare. Electroencephalography (EEG) signals monitor human brain activity, these signals are used to identify epileptic seizure diseases. A medical laboratory technician is necessary to measure the EEG signal and to differentiate normal signals and epileptic signals. However, this technique is time-consuming and costly. Accurately detecting seizures is of utmost importance in the administration of antiepileptic drug treatment to mitigate potential risks in the future. In this project, we use a Machine Learning (ML) algorithm namely XG-BOOST to predict seizures and classify them into four levels, pre-ictal, ictal, inter-ictal, and normal with their accuracy. The proposed system was able to achieve an accuracy of up to 96% for the real-time data set we extracted from the EEG sensor.