The time course and relation of positive signal-averaged electrocardiograms by time-domain and spectral temporal mapping analyses after infarction

Spectrotemporal mapping of high-resolution ECGs in experimental myocardial infarction: comparison with time-domain analysis and epicardial electrograms

The study was undertaken to evaluate the relationship of signal-averaged ECG (SA-ECG) readings in the frequency domain (STM) and epicardial electrograms (EE) recorded before and after acute myocardial infarction (AMI) in pigs and to compare the changes with findings in time-domain analysis (TDA). In 20 pigs the left anterior descending artery (LAD) was ligated. Prior to ligation, a SA-ECG was recorded (method of Simson) and bipolar electrodes were used to register EE in the areas supplied by the LAD and the circumflex artery (CIRC). Five minutes after LAD ligation, all measurements were repeated. Time-domain parameters were QRS duration (QRS D) and the duration of the signal below 30 microV (LAS 30). Beginning at a point of 20 ms before the QRS end, the frequency spectra (0-200 Hz) of 25 segments of 80-ms duration at the QRS end were analyzed. The volumes below the 25 curves were analyzed separately for 0-50 Hz, 51-100 Hz, 101-150 Hz, and 151-200 Hz. After AMI, five pigs died within 7 minutes. In 15 pigs, QRS D as well as LAS 30 lengthened significantly (P<0.05; P<0.001). Spectrotemporal mapping (STM) showed a significant decrease of the frequencies above 50 Hz (51-200 Hz) in the entire group and in the animals with late potentials (P<0.05). EE of the LAD area were significantly prolonged (P<0.01); this did not correlate with the changes in STM parameters. In pigs acute myocardial infarction causes a shift towards lower frequencies in the STM which most likely reflects the slowed depolarisation in the ischemic area.

High-resolution electrocardiography

High-resolution electrocardiography, often referred to as signal averaged electrocardiography (SAECG), has clinical utility in human medicine for detecting ventricular late potentials (LP) as predictors of future arrhythmic events, specifically reentrant type of ventricular tachycardia (VT). Time-domain analysis of the SAECG is conceptually similar to standard ECG analysis, but the primary objective is to detect LP, the marker for reentrant pathways within the myocardium. Frequency-domain analysis is investigational but provides the same information as time-domain analysis and both methods of analysis have limitations. The presence of LP has been associated with the ability to induce VT, unexplained syncope, and sudden arrhythmic death. Some cardiomyopathic dogs with VT have SAECG that appear to contain LP, and sudden death has occurred in these dogs.

Relationship between the site of a myocardial infarction and signal-averaged electrocardiogram indices

We evaluated the relationship between the site of a myocardial infarction (MI) and signal-averaged electrocardiogram (SAECG) indices in both time-domain (TDA) and spectral turbulence (STA) analyses, and their implications in the prediction of infarct-related artery (IRA) patency, in 114 survivors of a first MI. They were divided into two groups based on MI location (57 anterior and 57 inferior). Patients with bundle branch block were not included. Fifty patients had been treated with thrombolytic therapy (TT). The STA was done in both XYZ-leads and in vector magnitude. Forty patients had an abnormal SAECG in TDA and 37 in STA, but only 22 (19%) in both (71% of agreement, kappa=0.35). Fifty-four patients (47%) had an occluded IRA. The best predictors from multivariate analysis of having an occluded IRA in the inferior MI group were: an abnormal Y-lead in STA (odds ratio 4.9; P=0.005); an abnormal RMS40 in TDA (odds ratio, 4.8; P=0.02); absence of TT (odds ratio, 9.15; P=0.001). Conversely, in the anterior MI group they were: an abnormal SAECG in TDA (odds ratio 6.83; P=0.005); absence of TT (odds ratio, 4.3; P=0.02). The multivariate receiver operator characteristic curves clearly showed the effect of MI location on the SAECG indices. This study suggests that the myocardial infarction site is an important factor for the variability and poor concordance between TDA and STA. Such differences may alter the predictive accuracy of SAECG. TDA and STA should be complementary methods, and the exploration in each orthogonal lead appears to be better than in vector magnitude.

Frequency content and sex difference of the Frank lead signal-averaged ECG in a population with significant coronary artery disease. Comparison with concurrent 12-lead ECG morphology

The 12-lead electrocardiogram (ECG) and a 100-beat signal-averaged Frank lead ECG (SAECG) at a sampling rate of 1,000 Hz and with 16-bit resolution were recorded from 52 women and 256 men with significant coronary artery disease presenting for coronary artery revascularization. The QRS portion of each Frank lead was digitally filtered in four bandwidths: 0-10, 10-60, 60-150, and 150-250 Hz. The root-mean-square (RMS) voltage of each filtered signal was calculated as an absolute value and normalized as a percentage of the sum of the four filters, creating 27 variables. Three groups were formed using the presenting 12-lead ECG: N-ECG, ST-ECG, and MI-ECG. Despite variation in 12-lead morphology, concordance was dominant in the RMS values of the SAECG in sex comparisons within and between groups. There was significant sex difference of the RMS values in 6 of the 15 absolute RMS variables within the three groups. Women had no significant between-group difference, and men had a significant between-group differences in five absolute values and one normalized RMS value. P was considered significant at < .05.