Epileptic seizure detection based on the instantaneous area of analytic intrinsic mode functions of EEG signals

Abstract

Epileptic seizure is generated by abnormal synchronization of neurons of the cerebral cortex of the patients, which is commonly detected by electroencephalograph (EEG) signals. In this paper, the intracranial EEG signals have been used to detect focal temporal lobe epilepsy. This paper presents a new method based on empirical mode decomposition (EMD) of EEG signals for detection of epileptic seizures. The proposed method uses the Hilbert transformation of intrinsic mode functions (IMFs), obtained by EMD process that provides analytic signal representation of IMFs. The instantaneous area measured from the trace of the windowed analytic IMFs of EEG signals provides rules-based detection of focal temporal lobe epilepsy. The experiment results on intracranial EEG signals are included to show the effectiveness of the proposed method for detection of focal temporal lobe epilepsy. The performance evaluation of the proposed method for epileptic seizure detection has performed by computing the sensitivity (SEN), specificity (SPE), positive predictive value (PPV), negative predictive value (NPV) and error rate detection (ERD). The proposed method has been compared to the existing methods for detecting focal temporal lobe epilepsy from intracranial EEG signals. The proposed method has provided detection of focal temporal lobe epilepsy with increased accuracy.