T-wave detection using the empirical mode decomposition

Abstract

Identifying key elements of the ECG signal in modern day's diagnostic devices is crucial. Making a preliminary assessment of a patient during emergency situations, i.e. before arriving to the hospital, is starting to be a generally accepted practice in more and more countries. This article proposes a possible method of differentiating the T-wave from other features of the ECG signal based on the empirical mode decomposition. The method identifies the T-waves by removing the QRS-complexes from the signal and computing the intrinsic mode functions of this modified signal. These intrinsic mode functions offer a characteristic waveform containing the T-wave. An average positive predictibility of 95.99% is obtained.

Keywords

• ECG processing
• empirical mode decomposition
• intrinsic mode function
• Shannon energy
• T-wave