Assessment of the Similarity Between Vectorcardiogram Curves of Both Methods Frank Standard and Kors Regression Matrix Using Digital Image Processing

Abstract

The vectorcardiogram (VCG) is an useful method for automatic prognosis and diagnosis in different medical scenarios. However, its method of acquisition presents some disadvantages related to complexity of the standard system proposed by Ernest Frank. For this reason, Jan A. Kors proposed a method of reconstructing the VCG from the usual ECG leads. This paper studied the validation of the Kors’s technique by comparing the VCG plotted-by-plane images using two metrics of image processing. The group of healthy subjects presented the best results with mean PSNR of 18.137 dB and mean SSIM of 90.5% for VCG XY plane images.