Three-dimensional vectorcardiography (3-D VCG) by computer graphics in old myocardial infarction

Delineation of premature P waves on four-dimensional electrocardiography, a new display of electrical forces by computer techniques

This study investigated the feasibility of four-dimensional electrocardiography (4-D ECG), a new display in which the vector loop was rotated and scanned along a timed axis to overcome the shortcomings of vectorcardiography (VCG). The subjects consisted of 38 patients with premature atrial complexes and 30 controls. The orthogonal Frank electrocardiograms were rotated three-dimensionally according to the right-hand rectangular coordinate system and scanned along a timed axis. The P wave delineation score, signifying good agreement with the intraobserver and interobserver variability, was significantly higher in 4-D ECG than those in the orthogonal leads or those on the transverse and frontal projections (P < 0.001). The authors measured the premature P loop areas as viewed from 361 directions. P loop areas were best delineated when viewed from cranial directions of 42.6 +/- 34.0 degrees and from rightward directions of 11.3 +/- 30.7 degrees. Adequate cranial rotation followed by scanning along a timed axis will maximally delineate premature atrial signals and provide comprehensive visualization of electrical forces.

Analysis of bites on three-dimensional vectorcardiography after coronary artery ligation in dogs

The serial changes in the QRS loops of vectorcardiograms were investigated following ligation of a branch of the left anterior descending artery (LAD) by the three-dimensional rotation method in 18 dogs. Concave inflections of the QRS loop, defined as "bites," were best delineated when the loop was viewed from a left cranial or right caudal direction. Bites appeared 48 +/- 8 minutes after LAD ligation in all of the dogs, and their development was closely related to the temporal changes in the % sigma R and QRS point score on a standard 12-lead electrocardiogram. Q waves were not observed on the electrocardiograms in 9 dogs. In the remaining 9, they appeared 117 +/- 18 minutes after LAD ligation. The bite duration, area, and amplitude were compared with the anatomical extent of the infarcts. A significant positive correlation was found between bite duration and infarct size. The detection of bites on the three-dimensionally rotated vectorcardiogram appears to have a high sensitivity for anterior myocardial infarction and could potentially become a useful diagnostic tool.

Vectorcardiographic bites. A method for detection and quantification applied on a normal material

Vectorcardiographic bites as an expression of small fibrotic or necrotic areas in the myocardium have been discussed for many years. Distinct definitions of bites and normal limits for bites have, however, not been established. Therefore, an algorithm for computer detection and quantification of bites is presented. To find a bite, sectors of the QRS loop rotating opposite the way of the main part of the loop (eg, a clockwise-rotating sector in an otherwise counterclockwise-rotated loop) are detected. The bite is then delineated, using an iterative procedure. Finally the amplitude, duration, and area of the bite are calculated. The method for detection and quantification was applied on a well-defined normal material to obtain normal limits for bites. The commonly used criterion for an abnormal bite (amplitude greater than or equal to 0.1 mV and duration greater than or equal to 10 msec in the horizontal or sagittal plane) results in a specificity of only 87%. A specificity of 95% is yielded with the criterion of bite amplitude greater than 0.15 mV in the horizontal or sagittal plane.