Theoretical comments on reproducibility and normalization of TWA measures

Using a simple stochastic model of ventricular repolarization and the equivalent surface source (ESS) model, an electrophysiological formulation relating surface ECG to variations at the myocytes' level, we recently pointed out a few theoretical results regarding T-wave alternans (TWA). In this paper, stimulated by the comments of John E. Madias on our paper (J Electrocardiol, 2012), we further explored the consequences implied by the theoretical model. First, we verified the reproducibility of TWA measures, in clinically stable patients repeatedly tested. The sensitivity to displacement was evaluated simulating lead mislocations of up to 20mm. The numerical simulations were performed on data obtained solving the inverse electrocardiographically problem from three subjects (ECGSIM). The results showed that TWA sensitivity varies across leads, being maximal in V1 and decreases towards V6. Globally, the maximal percent error found was 6.1%. Thus, TWA measures do not seem to add more stringent requirements on lead placement's precision, than the usual diagnostic practice. Finally, we further discussed the implications of normalizing TWA measures. While clinical studies are necessary to sort out the issue, the theoretical model suggests that normalization might be appropriate only is certain cases.