Derivation of the 12-lead electrocardiogram and 3-lead vectorcardiogram

Abstract

ObjectiveThe cardiac dipolar field is represented by the measured 12-lead electrocardiogram (ECG) and 3-lead vectorcardiogram (VCG). The objective is to derive the 12-lead ECG and 3-lead VCG from 3 measured leads acquired from only 5 electrodes. MethodsThis is a retrospective blinded study comparing measured and derived ECG and VCG tracings. A nonlinear optimization model was used to synthesize the derived 12-lead ECG and 3-lead derived VCG from leads I, II, and V2. A total of 367 measured 12-lead electrocardiograms and 3-lead vectorcardiograms of varying morphologies were acquired from archived digital ECG databases. All tracings were interpreted by 2 blinded physician reference standards. The derived vs measured tracings were compared quantitatively using Pearson correlation and root mean square error. Qualitative comparisons were determined by physician percent agreement analysis and adjudication. ResultsThe correlations between the measured and derived ECGs and VCGs were high (r=0.867). No clinically significant differences were noted in 98.1% of cases. Electrocardiographic rate, rhythm, segment, axis, and acute myocardial infarction interpretations showed 100% correlation. Root mean square error compared favorably against other synthesis techniques. Overall percent agreements for the various ECG morphologies were noted to be 98.4% to 100%. ConclusionsThe 12-lead ECG and 3-lead VCG can be derived accurately from 3 measured leads with high quantitative and qualitative correlations. These derived tracings can be acquired instantaneously and displayed in real time from a cardiac rhythm monitor. This will allow for immediate, on-demand, convenient, and cost-effective acquisition and analysis of the 12-lead ECG and 3-lead VCG in areas of acute patient care.