Selected transformation methods and their comparison for VCG leads deriving

Abstract

Vectorcardiography (VCG) as an alternative form of 12-lead ECG is another method of measuring the electrical activity of the heart. The use of vectorcardiography in clinical practice is not common, but VCG leads can be derived from 12-lead ECG. VCG has proven to be a useful and more accurate tool for diagnosing various heart diseases within automated detection. This paper presents the application of four transformation methods namely: Kors regression, IDT, QLSV and Quasi-Orthogonal transformation to obtain a derived VCG. A total of 20 physiological and 20 records with the diagnosis of myocardial infarction were used. For physiological records, the Kors regression method achieved the best results in leads X and Y with relative deviation <1%, correlation and percentage similarity >99%. In lead Z, the QLSV method achieved the most accurate results with relative deviation <1%, correlation >98% and percentage similarity >99%. For pathological records, the most accurate method in all leads was Kors regression with relative deviation <2.2%, correlation >93% and percentage similarity >97%. From these results, there is the possibility of creating a new transformation method from the existing ones in order to obtain a more accurate transformation.