A new method of filtering T waves to detect hidden P waves in electrocardiogram signals

Abstract

A correct identification of the P wave is crucial for the diagnosis of narrow QRS tachycardias. This is sometimes difficult because atrial activity is hidden in the T wave. The aim of this study is to evaluate the usefulness of a T wave filtering technique based on wavelet transformation to identify atrial activity. Forty-two patients with narrow QRS tachycardias and regular atrial activity were studied. A surface electrocardiogram (ECG), intra-atrial recording, and the T wave filtering ECG were compared simultaneously to check the accuracy of the filtering system in detecting atrial activity. The sensitivity of the T wave filtering and P wave detection algorithm was 85.8% [95% confidence interval (CI): 81.2-89.4%] and the specificity was 89.4% (95% CI: 87.1-91.4%), with a global accuracy of 88.5% (95% CI: 86.5-90.3%). The expert cardiologist's accuracy in distinguishing between atrioventricular nodal reentry tachycardia and atrioventricular reentry tachycardia was 75% in the surface ECG vs. 100% in the ECG with the T wave filtering process (P<0.01). T wave filtering based on wavelet transformation improves the capacity of the surface ECG to identify atrial activity in cases of regular narrow QRS supraventricular tachycardias.