Optimizing ECG Signal Sampling Frequency for T-Wave Alternans Detection

Computer detection of microvolt T-wave alternans (TWA) is an non-invasive method to identify patients at high risk for ventricular arrhythmias. Since TWA is a transient phenomenon, there is the need for continuous long-term TWA analysis in Holter ECG recordings. TWA detection, usually detected in ECGs sampled at 1000 samples per second (sps), in computationally demanding. We determined the ability of our correlation method (CM) to identify TWA in ECGs sampled at lower frequencies. TWA was identified in 39 long QT syndrome patients, whose ECGs were originally acquired at 1000 sps, and then resampled at 100, 250, 500, and 750 sps. Results obtained at different sampling conditions were compared. We found that TWA can be effectively detected with the CM using sampling frequencies as low as 250 sps. Such sampling frequency seems to be optimal since it provides high accuracy of TWA measurements and substantial saving of computational time.