New algorithm for QT dispersion analysis in XYZ-lead Holter ECG. Performance and applications

In this paper a new algorithm for automatic analysis of QT dispersion (QTd) in three-lead (X,Y,Z) Holter ECG records is presented. To evaluate its performance the authors have compared automatic with manual annotations using a set of 1000 beats of Holter ECG records from the IDEAL Data Base. The signals have been automatically analyzed and the same beats have been analyzed by two cardiologists. The results obtained show that the reproducibility of the automatic QTd measurement is comparable to manual. QTd has been studied as a discriminator of patients with Dilative Cardiomyopathy from healthy subjects. The authors have found that QTd in 24 hours presents statistically significant differences when comparing subjects from those two groups. They have analyzed also mean and standard deviation of RR, QT and QTc intervals to determine their statistical significance in discriminating subjects from both groups.