The effect of baseline wandering in automatic T-wave alternans detection from holter recordings

T-wave alternans (TWA) is an electrophysiologic phenomenon associated with an increased risk of death. ECG baseline wandering may prevent correct detection of TWA. The present study was designed to test the effectiveness of our newly developed heart-rate adaptive match filter (AMF) to remove baseline wanders and improve TWA detection. To this aim, both simulated and experimental (10 Holter ECGs of patients with acute myocardial infarction, AMI) data were used. Performance of our AMF was compared with that of a third-order spline (TOS) interpolation. In simulated data, AMF allowed correct detection of TWA almost independently of baseline frequency components, when these were different from TWA own frequency, whereas the TOS interpolation allowed TWA detection only for baseline frequencies lower than TWA frequency. Application of AMF to AMI patients allowed detection of TWA in 2 out of 10, whereas 10 out of 10 were detected as TWA-positive after application of the TOS (p<0.05). The two patients that resulted TWA-positive after AMF application were characterized by a larger number (p<0.05) of beats involved in the TWA episodes than the number detected by the TOS. Detection of TWA in 100% of AMI patients provided by the TOS-technique (suggesting the presence of false-positive), and presence of false-negative, as deduced from the simulation results, question the reliability of this method. Our AMF allows more reliable identification of TWA almost independently of baseline frequency components.