T-wave Vector Alternans Detection based on Holter ECG Recordings

Abstract

This paper proposes a new method for detecting T-wave alternans (TWA) based on 3-channel Holter ECG recordings. The current standard method, based on spectral analysis of each signal lead, enables low amplitude alternans detection at the microvolt level. However, the method requires a controlled test environment where the mean heart rate is artificially increased. T Proposed method aims at realizing the reliable alternans detection from 24 hour Holter recordings during normal daily activities. To achieve this, the method utilizes singular value decomposition (SVD) for highly sensitive differentiation of T-wave morphology in noisy recording conditions. We propose the name T-wave Vector Alternance (TWVA) for TWA detected by SVD in decomposed ECG signals. The method was applied to three normal subjects and two subjects with TWA, and it correctly detected the TWA.