Ability of standard ECG parameters to detect the body surface isopotential abnormalities of pacing induced myocardial ischemia in the dog

Classification of pathologies by reduced sequential potential maps

Body surface potential mapping (BSPM) is an electrocardiographic measuring technique which produces the data as a series of three-dimensional maps. These maps are assumed to contain information which may help classify subjects for diagnostic purposes more effectively than standard ECGs. As quantitative classification of the complete sequences of maps is complex and cumbersome, the present study uses extracted features which characterise the data. The features, which have been presented and evaluated in a recent work, have been extracted after the maps were processed by a compression technique which conserved the spatial details of the maps. The compression by two-level thresholding converted the sequences of maps into sequences of annuli, from which the following features were extracted: time indices, velocity vector magnitude, loci in three-dimensional space of the centres of mass and cross-correlation coefficients between successive annuli in the sequence. Here, three different classification methods are applied to these features: statistical methods, the Fisher linear discriminant method and visual inspection. BSPMs from 54 subjects are used: 25 normal, 11 WPW syndrome and 18 CAD cases. It is found that by applying a decision role which comprises all features, the procedure offers a completely accurate classification of the subjects to their groups. The three-dimensional centre of mass is found to be the single best classifier; successfully categorising 20/25 of the normals 17/18 of the CAD patients and 11/11 of the WPW patients.

Conservation and characterisation of spatial features in a new method of data compression for body surface potential maps

Body surface potential maps consist of a huge amount of data represented as a series of three-dimensional maps, which are time consuming to process and expensive to store. In spite of the continuous interest in body surface potential maps, their use has not become common and they are of no practical use in the clinics. This is due to the overwhelming amount of measured data required to generate the maps and the lack of quantitative methods to analyse them. Data compression or reduction may solve these deficiencies. Such a procedure must conserve the fine spatial details of the maps, which are usually extracted from low level surface potentials, as these are reported to be significant in diagnostic electrocardiography. A technique is presented for data reduction, that implements two-level thresholding and conserves the fine significant spatial features of each map. A sequence of annuli thus produced is shown to describe the dynamic nature of the underlying process. This sequence is further processed and characterised by features which quantify its dynamic behaviour: time of annuli sequence appearance, its duration, three-dimensional loci of centres of mass of the annuli, distances between successive centres of mass and cross-correlation coefficients between successive annuli. To test the data reduction procedure and the usefulness of the features, maps from 20 subjects are studied (both normal patients and those with various pathologies). It is found that the use of annuli instead of the whole measured information allows simple storage, display and calculations; the features, which vary in time, represent closely the changes in location of the annuli and their dynamic variations of shape. The features are also found to be grouped together for the maps of the normal patients and for each pathology. Thus, body surface potential maps may become more commonly used in clinics by being represented by a set of features, which conserve their dynamic and spatial nature, and which may serve for classification of cardiac pathologies.

Epicardial electrical activation analyzed via frequency-wavenumber spectrum estimation for the characterization of arrhythmiagenic states

The application of a new signal processing methodology to the analysis of epicardial array ECG signals is presented as an alternative to isopotential or isochrones mapping by the use of a zero-delay wavenumber spectrum (ZDWS) estimation technique. The methodology "explains" the array data as the sum of modulated wideband (non-sinusoidal) propagating waves projected onto the array plane and provides an accurate estimate of their number and bearing. The slowness distribution of each of the waves is then obtained by estimating their temporal spectrum. In this experimental study the effects of localized noninfarcting reversible low flow ischemia, digoxin toxicity and verapamil reversal of digoxin toxicity are quantified via the ZDWS methodology and are compared with the information that can be extracted from isopotential mapping. It is demonstrated that the ZDWS methodology permits the epicardial electric activation to be decomposed into a number of quantification parameters which possess a hierarchical "tree" structure and therefore provide a means for an objective and robust characterization of the effects of agents which alter myocardial conduction and arrhythmia generation.

Alterations in transmural blood flow and body surface ST segment abnormalities produced by ischemia in the circumflex and left anterior descending coronary arterial beds of the dog

Previous studies have documented a quantitative relation between alterations in transmural myocardial blood flow and body surface electrocardiographic distributions during rapid atrial pacing after chronic occlusion of the left circumflex coronary artery (LCx). Because other studies have described functional differences between the left anterior descending (LAD) and the LCx perfusion beds, we tested the hypothesis that these two territories exhibit quantitative differences in their responses to demand-dependent myocardial ischemia. To do so, 25 sedated dogs were studied 3 weeks after implantation of an ameroid constrictor around the proximal LCx (15 dogs, group I) or the LAD (group II). Oxygen demand was increased by rapid atrial pacing at rates of 90 to 210 beats/min, myocardial blood flow was measured by serial injections of radiolabeled microspheres, and the electrocardiographic consequences were evaluated by isopotential body surface mapping. Endocardial flows and the endocardial/epicardial flow ratio fell to significantly lower levels during atrial pacing in the ischemic LAD bed than in the LCx perfusion zone. Electrocardiographic patterns indicative of subendocardial ischemia also developed with lesser abnormalities in endocardial/epicardial ratios as determined by logistic regression models, in the LAD than in the LCx bed. Thus the LAD bed is more susceptible to ischemia than the LCx region because of differences in collateral blood flow patterns. In addition, the intensity of the surface electrocardiographic potentials during ischemia was significantly greater, as measured by linear regression, after LAD than after LCx obstruction. These data thus demonstrate significant differences between the two cardiac regions as electrocardiographic potential sources during ischemia.

Regional blood flow correlates of ST segment depression in tachycardia-induced myocardial ischemia

Tachycardia produces subendocardial ischemia and ST segment abnormalities after coronary obstruction. To determine whether a quantitative relationship exists between these ST shifts and transmural blood flow, 19 dogs were studied. Coronary obstruction was produced by ameroid constriction of the left circumflex artery, and tachycardia was generated by atrial pacing at 90 to 210 beats/min. ST shifts were studied by body surface isopotential mapping with an 84-electrode torso grid, and blood flow was quantitated by serial radiolabeled microsphere injections. Isopotential maps at each paced rate, 40 msec into the ST segment, were classified as normal or ischemic based on spatial patterns of voltages. Pacing after 3 weeks of ameroid constriction reduced endocardial/epicardial flow ratios in 11 dogs from 1.16 +/- 0.22 at rest to 0.41 +/- 0.18 at 210 beats/min. Abnormal ST depression developed in these dogs at a rate of 184.0 +/- 16.5 beats/min. Endocardial/epicardial ratios with ST depression (0.45 +/- 0.15) were lower than at those without ST depression (1.05 +/- 0.19; p less than .01). Logistic regression analysis demonstrated that ST depression corresponded to an endocardial/epicardial ratio of 0.67 or less (p less than .01). With this model, 95.5% of data sets were correctly classified. Neither heart rate nor perfusion bed size were significant independent predictors of an ischemic electrocardiographic response. The magnitude of abnormal ST segment shift was significantly correlated (r = .87) with the transmural flow ratio. Thus development of electrocardiographic changes indicative of ischemia corresponds to a predictable degree of flow redistribution and the magnitude of the ST shift is correlated with the intensity of the flow abnormality.