Open AccessAnalysis of EEG signals using Machine Learning for the Detection and Diagnosis of Epilepsy
Author(s) -
Anubha Nagar,
Bidushi,
Mimangsha Sarma,
Mithra Anand Kumar,
J. Valarmathi
Publication year - 2020
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.a1685.1010120
Subject(s) - epilepsy , electroencephalography , computer science , artificial intelligence , kurtosis , decision tree , skewness , machine learning , epileptic seizure , logistic regression , sensitivity (control systems) , pattern recognition (psychology) , feature extraction , set (abstract data type) , random forest , statistics , psychology , mathematics , psychiatry , engineering , electronic engineering , programming language
Electroencephalogram (EEG) is one of the most commonly used tools for epilepsy detection. In this paper we have presented two methods for the diagnosis of epilepsy using machine learning techniques. EEG waveforms have five different kinds of frequency bands. Out of which only two namely theta and gamma bands carry epileptic seizure information. Our model determines the statistical features like mean, variance, maximum, minimum, kurtosis, and skewness from the raw data set. This reduces the mathematical complexities and time consumption of the feature extraction method. It then uses a Logistic regression model and decision tree model to classify whether a person is epileptic or not. After the implementation of the machine learning models, parameters like accuracy, sensitivity, and recall have been found. The results for the same are analyzed in detail in this paper. Epileptic seizures cause severe damage to the brain which affects the health of a person. Our key objective from this paper is to help in the early prediction and detection of epilepsy so that preventive interventions can be provided and precautionary measures are taken to prevent the patient from suffering any severe damage