Optimal solution to patient-specific coefficients for electrocardiogram to vectorcardiogram conversion using iterative multiple linear regression

Eduardo Morales, University of Texas at El Paso

Abstract

Vectorcardiography, the study of the electrical activity of the human heart in vector form, has not been fully applied to clinical practice due to patient-specificity. Considering that the only existing methods for measuring the vectorcardiogram are based on ideal human models, vectorcardiography has not been considered clinically practical because existing methods do not take into account and ignore patient-specific properties. ^ In this research, it is shown that using the mentioned method provides an accuracy of 5.66% average rms error (ARMSE) and a correlation coefficient of 0.97 for 10 normal patients as compared to 14.58% ARMSE and a correlation coefficient of 0.86 for a method available in the literature (Dower et al.). Plus, a 7.13% ARMSE and a correlation coefficient of 0.95 were found for 10 abnormal patients as compared with 20.74% ARMSE and a correlation coefficient of 0.78 for Dower et al. parameters. The improvement of ARMSE and correlation coefficient obtained from the proposed method suggest its application to clinical practice. (Abstract shortened by UMI.) ^

Subject Area

Engineering, Biomedical|Engineering, Electronics and Electrical

Recommended Citation

Morales, Eduardo, "Optimal solution to patient-specific coefficients for electrocardiogram to vectorcardiogram conversion using iterative multiple linear regression" (2005). ETD Collection for University of Texas, El Paso. AAI1430227.
http://digitalcommons.utep.edu/dissertations/AAI1430227

Share

COinS