Continuous time analysis method for T-wave alternans detection

T-wave alternans (TWA) or repolarization alternans have been closely linked to electrical instability of the heart. Recent studies have demonstrated that TWA ocurrence is related with ventricular arrhythmias. Accordingly, TWA detection may help in the diagnosis of heart diseases and sudden cardiac death. It consists of every-other-beat subtle fluctuations in the repolarization section ranged in the order of μV. For such small variations, noise is a critical factor. Noise amplitudes similar to the alternant wave may make TWA detection difficult. In this work, a new time domain method for TWA detection is presented. The performance evaluation is carried out with simulated signals through comparative studies against the well known Spectral Method (SM). The proposed method yields better performance in a noisy environment with a decreased computational cost, and is suitable for embedded real time implementation.