Whole brain epileptic seizure detection using unsupervised classification

Abstract

Epilepsy is one of the most prevalent neurological disorders in human beings. It is characterized by recurring seizures in which abnormal electrical activity in the brain causes the loss of consciousness or a whole body convulsion. The seizure detection is an important component in the diagnosis of epilepsy to figure out the causes, mechanisms and treatment. In the clinical practice, this detection involves visual scanning of Electroencephalogram (EEG) by the epileptologist in order to detect and classify the seizure activity present in the EEG signal. Automated detection of correlates of seizure activity across all regions of the brain and across can be a solution. This paper proposes New framework of automatic Whole brain epileptic event detection using fast potential-based hierarchical agglomerative (PHA) Clustering Method and Empirical Mode Decomposition (EMD). Different distance such as Euclidian, Batacharay and kolomogorov were computed between the IMFs and used as input for the PHA cluster. The evaluation results are very promising indicating an overall accuracy of 98.84%.