Gilberto Sierra

Predictive Value of Wavelet Correlation Functions of Signal-Averaged Electrocardiogram in Patients After Anterior Versus Inferior Myocardial Infarction

OBJECTIVES This study sought to evaluate the prognostic value of wavelet correlation functions of the signal-averaged electrocardiogram (ECG) for arrhythmic events in patients after myocardial infarction. BACKGROUND Wavelet transform of the signal-averaged ECG has been shown to be a nonstationary analysis technique describing the time evolution of frequency spectra throughout the QRS complex. To quantify the wavelet transform, we introduced the new concept of the wavelet correlation function. METHODS The relation among wavelet correlation functions, ventricular late potentials and the site of infarction was investigated in 769 men < 66 years old who survived the acute phase of myocardial infarction (351 [46%] anterior, 418 [54%] inferior infarctions). Signal-averaged ECG recordings were obtained 2 to 3 weeks after infarction. During 6 months of follow-up, 33 patients (4.3%) experienced a malignant arrhythmic event. Wavelet correlation functions of the signal-averaged ECG were evaluated in a time-frequency plane ranging from 25 ms before QRS onset to 25 ms after QRS offset in the frequency range between 40 and 100 Hz. RESULTS Patients with an anterior infarction had lower mean wavelet correlation coefficients (p < 0.001) and a lower incidence of ventricular late potentials than patients with an inferior infarction (32.3% vs. 42.7%, p = 0.003). The combination of wavelet correlation functions and late potentials increased the total predictive accuracy from 52% to 72% for inferior and from 64% to 76% for anterior infarctions. CONCLUSIONS Spectral changes in the signal-averaged QRS complex are more prominent in anterior than inferior infarctions. Combination of late potential analysis and wavelet correlation functions increases the prognostic value for serious arrhythmic events after myocardial infarction.

Noninvasive Risk Modeling After Myocardial Infarction

The aim of this study was to extract and combine non-invasive risk parameters from the signal-averaged electrocardiogram (SAECG) and heart rate variability (HRV) based on 24-hour ambulatory electrocardiography to optimize the prognostic value for arrhythmic events after acute myocardial infarction. A prospective series of 553 men < 66 years of age enrolled in the Post-Infarction Late Potential study were analyzed. Within 2 to 4 weeks after acute myocardial infarction, all patients underwent SAECG and 24-hour ambulatory electrocardiography before hospital discharge. During 6 months of followup, 25 patients (4.5%) experienced arrhythmic events (sustained ventricular tachycardia, n = 11; ventricular fibrillation, n = 7; sudden cardiac death, n = 7). The predictive power of SAECG and HRV parameters was assessed using a Cox proportional-hazards model. In HRV analysis, the most significant differences between patients with and without arrhythmic events were observed for the beat-to-beat parameter root-meansquare of successive RR differences [RMSSD]): 25.7 +/- 16.9 ms in patients with arrhythmic events versus 34.1 +/- 18.6 ms in patients free of arrhythmic events (p = 0.004). Time domain analysis of the SAECG showed the QRS duration to be most significantly different in both patient groups: 106.4 +/- 18.7 ms (arrhythmic events) versus 95.3 +/- 18.7 ms (no arrhythmic events) (p = 0.001). Based on the Cox regression model, RMSSD and QRS duration were demonstrated to be independent significant risk factors (regression coefficient for QRS duration: cq = 0.014 +/- 0.006 ms(-1), p = 0.014; for RMSSD: cr = -0.041 +/- 0.016 ms(-1), p = 0.009). Based on the regression coefficients, an analytic risk model was developed describing the arrhythmic risk as a function of QRS duration, RMSSD, and time after infarction. We conclude that the combination of beat-to-beat changes of heart rate measured by RMSSD and QRS duration from the SAECG enhances noninvasive risk stratification after myocardial infarction.

Multiresolution decomposition of the signal-averaged ECG using the mallat approach for prediction of arrhythmic events after myocardial infarction

The aim of this study was to analyze the ability of the multiresolution decomposition of the signal-averaged electrocardiogram (ECG) to discriminate between patients who develop life-threatening ventricular arrhythmias after myocardial infarction and those who do not and to compare the predictive values of this approach with those obtained from the analysis of ventricular late potentials in the time domain. Signal-averaged ECGs of 769 prospectively included patients were analyzed. A total of 42 arrhythmic events occurred during the follow-up period. For numerical calculations of wavelet analysis, the total and relative energies of the QRS complex were obtained in seven frequency bands. The combination of the relative energy in the frequency bands 7.8-15.6 Hz and 62.5-125 Hz enhanced statistical performance as compared with the time-domain parameters (positive predictive accuracy, 11.3 vs 8.2%). Combining wavelet transform and time-domain parameters enhanced the predictive values even more (positive predictive accuracy, 14.3%) compared with applying each method alone.

Fluoroscopic navigation to guide RF catheter ablation of cardiac arrhythmias

Severe disorders of the heart rhythm that can cause syncope or sudden cardiac death (SCD), can be treated by radio-frequency (RF) catheter ablation. The precise localization of the arrhythmogenic site and the positioning of the RF catheter over that site are problematic: they can impair the efficiency of the procedure and are time consuming (several hours). Our approach consists of integrating fluoroscopic and electrical data from the RF catheters into the same image so as to better guide RF ablation, shorten the duration of this procedure and increase its efficacy.

Cycle Length Variability and Tachycardia Sites of Origin in a Canine Model

The purpose of this study was to examine the spontaneous changes in cycle length during episodes of sustained monomorphc (MVT) and polymorphic (PVT) ventricular tachycardas and to relate these changes with the sites of earliest epicardial activation. Isochronal activation maps were obtained from 127 unipolar electrograms recorded from the surface of both ventricles with a sock electrode array in open chest anesthetized dogs. Cycle length (431 * 80 msec) in MVT (7 dogs) had beat to beat variations of 15 f 17 msec corresponding with small shfts in sites of epicardial breakthrough clustered along the border of the ischemic myocardium. In PVT (5 dogs), the cycle length (352 k 90 msec) had a variability of 61 k 23 msec, corresponding to wide changes on the sites of epicardial breakthrough. Mean cycle length and variability were significantly different @<O.O 1) between MVT and PVT. Linear regression analysis showed a strong and significant correlation between cycle length variability and the number of electrodes detecting the earliest activation (~0.77, p<O.OOOl).

Intracoronary thrombolysis in patients with ST-segment elevation myocardial infarction presenting with massive intraluminal thrombus and failed aspiration

Aim: Massive intracoronary thrombus is associated with adverse procedural results including failed aspiration and unfavourable reperfusion. In this scenario the best treatment remains unknown. We aim to evaluate the effect of low dose intracoronary thrombolysis in patients with ST-segment elevation myocardial infarction (STEMI) presenting with a large thrombus burden and failed aspiration. Methods and results: In this study 30 STEMI patients with a large thrombus burden and failed manual aspiration were included in the registry. Local fibrinolysis with low dose (one-third of systemic) tenecteplase (27%) or alteplase (73%) was administered via a 6F-infusion catheter. A thrombus was qualified as large when its maximal dimension was ≥2 vessels (thrombolysis in myocardial infarction (TIMI) thrombus grade 4 and 5). Altogether 33% of patients received delayed invasive therapy (>12 h) guided by the presence of recurrent symptoms and/or ischaemia. A total of 17% of patients presented with cardiogenic shock and half of the population received glycoprotein IIb/IIIa inhibitors. A post-lysis thrombus grade 4–5 was found in just one patient. TIMI flow grade improved from 0/1 at baseline (93%) to ≥2 in most patients (97%). Blush grade 2–3 was observed in 85% and we observed a similar percentage (82%) of more than 50% ST-segment resolution. In-hospital mortality was 10% (three patients with cardiogenic shock). No major bleeding events were observed. At a median follow-up of 14 (6–35) months, only one new clinical event was reported (1 target lesion revascularization, (TLR)). Conclusions: In STEMI patients with a large thrombus burden and failed manual aspiration, administration of low dose intracoronary thrombolysis is safe and reduces trhombus burden, as a result improving in epicardial flow and myocardial reperfusion.

The coherence spectrum in an experimental model of ventricular fibrillation

The present paper investigated the use of the coherence spectrum as a method able to discriminate ventricular fibrillation and nonfibrillatory rhythms, when applied to the surface ECG in an animal experiment using rabbits. The magnitude-squared coherence function was calculated using three and five segments averaged and compared with a threshold; if it is greater than the threshold it is a nonfibrillatory rhythm otherwise it is ventricular fibrillation. The sensitivity and specificity obtained in this study encourage the extension of this technique to the analysis of the surface ECGs in humans.

Application of learning vector quantization for localization of myocardial infarction

In this study myocardial infarction was localised by a Learning Vector Quantization (LVQ) classifier. Only information about ST-elevations in all 12 leads of the standard ECG were used. The significance of proper initialisation is demonstrated. A total classification accuracy of 85.6% was achieved by a classifier trained with the optimized-learning rate LVQ1 and 50% of the 769 patients. When the classifier was further trained with the LVQ2.1 and the LVQ3 algorithms no significant improvement in the classification accuracy was observed.

Discrimination Between Monomorphic and Polymorphic Ventricular Tachycardia Using Cycle Length Variability Measured by Wavelet Transform Analysis

Abstract The objective of this study was to assess the capability of wavelet transform (WT) analysis to differentiate between monomorphic (MVTs) and polymorphic ventricular tachycardias (PVTs) in a canine model and to relate these results to epicardial isochronal maps on a beat-by-beat basis. Unipolar electrograms were simultaneously recorded from the surface of both ventricles with a 127-lead sock electrode array in 24 open-chest anesthetized dogs. The sampling frequency was 500 Hz. Atrioventricular block was induced by formaldehyde injection into the atrioventricular node. The left anterior descending coronary artery was occluded for 60 minutes under ventricular pacing (140 stimuli/min) followed by reperfusion. Ventricular tachycardias were obtained during reperfusion and during left stellate ganglion stimulation. After visual selection, a total of 97 segments of 2,048 samples (4.096 seconds) were extracted and classified as 67 MVTs and 30 PVTs. A parameter based on the cycle length variability was defined in the second scale of the WT decomposition, normalized by its mean value. Similar assessment of cycle length variability was performed based on the detection of the point of most rapid change in potential with a negative slope in excess of −0.5 mV/ms in each individual electrogram to test the accuracy of the results obtained with the WT parameter. The WT parameter correctly identified 97% MVT and 83.3% PVT segments, for an overall accuracy of 92.8%. Beat-by-beat epicardial maps of MVT displayed a cluster of sites of initial activation close to the reperfusion area, while the sites of breakthrough from beats during PVT were much more dispersed over both ventricles. A strong and significant correlation was found between the number of electrodes with the earliest epicardial activation and the WT parameter ( r = .78, P P = .0022), and a strong linear correlation was found between both measurements (Pearson correlation coefficient, .966). It is concluded that WT analysis discriminated accurately between MVT and PVT, and a quantitative relation was found with the spatial dispersion of sites of earliest epicardial activation. The WT results strongly correlated with those obtained by another method of estimating cycle length variability. Methodologically, the strength of the WT lies in the complementary information that could be extracted from the processing of electrograms to enhance the detection/discrimination of different types of arrhythmias.

Multiresolution decomposition of the signal averaged ECG in patients with myocardial infarction compared to a control group

Wavelet transformation based on multiresolution decomposition using the method proposed by Mallat (1989) was applied to the signal-averaged ECG in patients with myocardial infarction and in a control group including healthy volunteers. The total and relative power of the QRS complex was calculated in seven frequency bands: 125-250 Hz, 62.5-125 Hz, 31.25-62.5 Hz, 15.62-31.25 Hz, 7.8-15.62 Hz, 3.9-7.8 Hz and below 3.9 Hz. Significant differences (p<0.0001) were found in all bands. To assess the discriminant potential of multiresolution decomposition in both groups, the sensitivity and specificity were calculated for each band. In the band from 125-250 Hz the highest sensitivity and specificity were found (88.5% and 66% respectively). The wavelet transform allowed a distinction between both groups and represents a promising processing tool in the field of ECG analysis.