Multi-parameter analysis of electroencephalogram (EEG): A diagnostic measure for epilepsy

Epilepsy is one of the most common neurological disorders and affects almost 60 million people worldwide. Epileptic seizures, which are characterized by a recurrent and sudden malfunction of the brain, reflect the clinical sign of an excessive and hyper-synchronous activity of neurons in the brain. EEG is the most economical one with high temporal resolution. The visual inspection of EEG data is prohibitively time-consuming and inefficient, even if the expert clinician reads the data ten times faster than the recording speed. The visual inspection lacks quantitative analysis which can uncover hidden characters of the data. We demonstrate in this paper with interictal scalp EEG data, which is much easier to collect than the ictal data, to automatically diagnose whether a person is epileptic or not. By using multi-parameter quantification and signal analysis such as power spectrum, frequency, amplitude, signal to noise ratio, entropy etc., it is possible to detect the epilepsy with affected lobe for localization. The results are promising with 100% sensitivity and accuracy of 91 % is achieved.