作者: Daryl G. Beetner , R. Martin Arthur
DOI: 10.1114/1.223
关键词:
摘要: New methods, based on Tikhonov regularization, were developed to infer the magnitude and phase of pericardial potentials directly. These methods tested in an adult-male torso model using measured human epicardial potentials. With 1% noise added body-surface potentials, regularization with optimal parameter at each frequency from 1 100 Hz gave average relative error (RE) inferred spectral magnitudes 0.44. Regularization composite–residual–smoothing–operator (CRESO) increased RE slightly 0.47. 10% additive noise, 10 mm overestimation heart radius, a position, CRESO 0.71. Performance was dependent. The smallest REs occurred low frequencies. 0.20, 0.40, 0.53 1–15, 15–46, 46–100 bands, respectively. Direct inference more accurate than Fourier transformation time-domain waveforms. Results suggest that when size location are not known, minimum estimates found overestimated which is value over band interest. © 1999 Biomedical Engineering Society. PAC99: 8719Hh, 8719Nn, 0260Lj