In order to elucidate the diagnostic implication of human body surface cardiac potential distribu tion, a computational reconstruction method for the body surface isopotential map (hereafter abbreviated as map) was developed by using the results of simulated ventricular propagation process, and the transfer impedance vectors which were actually measured with a human torso model.
For determining the values of the transfer impedance vectors in vast numbers, an automated processing system was designed, which was composed of a switchbox, amplifiers, a multiplexer, A/D and D/A converters, a minicomputer (core memory : 4 kw), a monitorscope and output devices.
Eighty-one electrodes implanted in the torso model were connected to the 9-channel switchbox. As often as an artificial dipole was successively moved on 392 positions which cover a ventricular area in the torso model, it was energized by a unit of sinusoidal current (100 Hz, 3 mA) in the three orthogonal directions and then the potential differences between each of 81 electrodes and Wilson's central terminal were fed to the switchbox. By switching manually, every nine out of 81 potential differences was simultaneously amplified, and scanned by the multiplexer. These analog-signals were A/D converted with a sampling rate of 2000/sec, and summed up with theminicomputer by an on-line procedure. These digitized values were D/A converted to be displayed on the monitorscope for the verification of the data. If found acceptable, these values were printed by teletype and punched out on paper tapes. In, such a way, about hundred thousand values of transfer impedance vectors were measured.
Utilizing these values, 81 potential values on the torso at each stage of the simulated ventricularpropagation process were calculated mathematically with a large computer, and the maps were drawn automatically by an X-Y plotter.
When these reconstructed maps were compared with those of normal human subject, the result showed that both map patterns were similar to each other. Therefore, it was recognized that this, reconstruction method for the maps was appropriate for most parts.
In relationship between ventricular activation and the maps, main wavefronts in the ventricular model at each stage were reflected properly on the reconstructed maps. Consequently, it is considered that the maps represent the ventricular activation more spatially and visually than conventional electro-and vector-cardiograms.
