ECG variability contour--a reference for evaluating the significance of amplitude ECG changes in two states.

The sensitivity of the standard 12-lead ECG for detecting myocardial ischemia is low perhaps because in the printed ECG only three to five complexes are examined, only absolute amplitude (voltage) criteria are used, and amplitude changes are sought only in the ST segment and T wave regions. A computerized method is proposed for evaluating the significance of ECG amplitude changes detected in one state compared with another (e.g. rest and stress). Amplitude changes were considered significant if they were consistently greater than the 'ECG variability contour' (EVC), which is a graphic measure introduced in this study, calculated from the reference ECG signal. Rest and stress simulation ECG (SECG) were constructed. Mean rest SECG complex was subtracted from rest and stress SECG complexes to result in rest and stress residue matrices, respectively. The percentage of the normalized cumulative sum (NCS) of the residues during stress (and rest) lying outside the EVC served as a measure for evaluating ECG changes associated with stress (and rest). With this method, amplitude changes of magnitude similar to that of the noise, which were difficult to detect visually, were easily detected and accurately allocated to the component of the ECG complex where they occurred. The proposed method may be useful in cases where amplitude changes are too subtle and thus overlooked or not detected by the standard examination of three to five complexes or underestimated due to unmet clinical (voltage) criteria, or occur in ECG components that are not regularly examined.