Toward T-Wave Recognition of ECG Signals Through Modulated Ensemble Empirical Mode Decomposition

Abstract

The cardiovascular diseases are the major cause of death globally. To diagnose heart disease, automatic recognition of ECG’s T-wave is necessary. Empirical mode decomposition (EMD) can be used to decompose nonlinear and nonstationary signals. However, using EMD to decompose ECG potentially leads to a mode mixing problem. This study proposes modulated EEMD (mEEMD) as a solution, which can solve mode mixing problems with almost no influence from noise. Furthermore, the mEEMD has a less problematic boundary side effect and does not cause any phase shift. The sensitivity of T-wave onset and offset recognition is [Formula: see text] and [Formula: see text].