Revista Cubana de Investigaciones Biomédicas (Jul 2020)
Ortogonalization of electrocardiogram’s leads
Abstract
Introduction: The Electrocardiographic (ECG)Signal Multiderivation Delineator based on the Wavelet Transform has high spatial resolution and allows eliminating interderivation differences that traditionally appear in uniderivation methods. For this, it need orthogonal ECG leads to each other to obtain a space loop. Objectives: To develop orthogonalization methods of two or three ECG leads that allow the generalization of the multiderivation delineator based on the wavelet transform in any database with more than one derivation. Methods: Three methods of orthogonalization of ECG leads were implemented: orthogonalization of two leads from vector projection, orthogonalization from main components and orthogonalization from the classical Gram-Schmidt method. Results: The operation of the ECGmultiderivation delineator was compared when each orthogonalization method is used, by calculating the arithmetic mean and the standard deviation taking into account different combinations of derivations of both databases for each of the brands analyzed. The best results were obtained with the Principal Component Analysis method and the worst behavior with the two lead orthogonalization method. Conclusions: The orthogonalization algorithms that obtained the best results were those based on three orthogonal leads, being slightly higher the decomposition into main components, and therefore it is considered the most suitable method for generalization of the multiderivationdelineator.