Seizure Classification Using Parallel Genetic Naive Bayes Classifiers

Epilepsy affects 50 million people worldwide and is one of the most common serious brain disorders. Seizure detection and classification is a valuable tool for maintaining the condition. An automated detection algorithm will allow for accurate diagnosis. This study proposes a method using unique features with a novel parallel classifier trained using a genetic algorithm. Ictal states from the EEG are segmented into 1.8 s windows, where the epochs are then further decomposed into 13 different features from the first IMF. All of the features are fed into a genetic algorithm (Binary Grey Wolf Optimisation Option 1) with a Naive Bayes classifier. Combining the simple partial and complex partial seizures provides the highest accuracy of all the models tested.