A Comparative Investigation of Mode Mixing in EEG Decomposition Using EMD, EEMD and M-EMD
Raymond H. Ho, Kevin Hung
Abstract
This paper presents and discusses the empirical mode decomposition method (EMD), its mode mixing problem and the primary causes of it. The ensemble EMD (EEMD) and masking EMD (M-EMD) methods are discussed as solutions for the mode mixing problem. Both improved methods were applied to process an EEG signal and the results were compared. The M-EMD provided the best result making good separation of the EEG signal into delta, theta, alpha, and beta band signals.
Topics & Concepts
Hilbert–Huang transformMixing (physics)ElectroencephalographyArtificial intelligenceSIGNAL (programming language)Pattern recognition (psychology)Computer scienceSpeech recognitionMode (computer interface)Signal processingMasking (illustration)PhysicsDigital signal processingComputer visionPsychiatryProgramming languageFilter (signal processing)PsychologyVisual artsQuantum mechanicsOperating systemArtComputer hardwareBlind Source Separation TechniquesMachine Fault Diagnosis TechniquesEEG and Brain-Computer Interfaces