Phase and amplitude coupling feature extraction and recognition of Ictal EEG using VMD

Automatic classification of epileptic signals plays a critical role in long term monitoring and diagnosis. This study provides a novel feature extraction and pattern recognition technique for classifying epilepsy using phase-phase, phase-amplitude and amplitude-amplitude coupling based on Variational Mode Decomposition (VMD). The EEG signal was decomposed to band limited intrinsic mode functions (BLIMFs) in the first place. Second, a coupling of BLIMFs was extracted as a feature and after determining an optimal feature two classifiers, strictly speaking, Random Forest and Extreme Learning Machine have classified the inputs. In order to evaluate and compare the performance of the new presented method Bonn University dataset of epilepsy have been used. Results indicate a more desirable performance than any other applied method suggested previously.