Effect of anisotropy on ventricular vulnerability to unidirectional block and reentry by single premature stimulation during normal sinus rhythm in rat heart.

Single high-intensity premature stimuli when applied to the ventricles during ventricular drive of an ectopic site, as in Winfree's "pinwheel experiment," usually induce reentry arrhythmias in the normal heart, while single low-intensity stimuli barely do. Yet ventricular arrhythmia vulnerability during normal sinus rhythm remains largely unexplored. With a view to define the role of anisotropy on ventricular vulnerability to unidirectional conduction block and reentry, we revisited the pinwheel experiment with reduced constraints in the in situ rat heart. New features included single premature stimulation during normal sinus rhythm, stimulation and unipolar potential mapping from the same high-resolution epicardial electrode array, and progressive increase in stimulation strength and prematurity from diastolic threshold until arrhythmia induction. Measurements were performed with 1-ms cathodal stimuli at multiple test sites (n = 26) in seven rats. Stimulus-induced virtual electrode polarization during sinus beat recovery phase influenced premature ventricular responses. Specifically, gradual increase in stimulus strength and prematurity progressively induced make, break, and graded-response stimulation mechanisms. Hence unidirectional conduction block occurred as follows: 1) along fiber direction, on right and left ventricular free walls (n = 23), initiating figure-eight reentry (n = 17) and tachycardia (n = 12), and 2) across fiber direction, on lower interventricular septum (n = 3), initiating spiral wave reentry (n = 2) and tachycardia (n = 1). Critical time window (55.1 ± 4.7 ms, 68.2 ± 6.0 ms) and stimulus strength lower limit (4.9 ± 0.6 mA) defined vulnerability to reentry. A novel finding of this study was that ventricular tachycardia evolves and is maintained by episodes of scroll-like wave and focal activation couplets. We also found that single low-intensity premature stimuli can induce repetitive ventricular response (n = 13) characterized by focal activations.NEW & NOTEWORTHY We performed ventricular cathodal point stimulation during sinus rhythm by progressively increasing stimulus strength and prematurity. Virtual electrode polarization and recovery gradient progressively induced make, break, and graded-response stimulation mechanisms. Unidirectional conduction block occurred along or across fiber direction, initiating figure-eight or spiral wave reentry, respectively, and tachycardia sustained by scroll wave and focal activations.

Myocardial electrical propagation in patients with idiopathic dilated cardiomyopathy.

Myocardial propagation may contribute to fatal arrhythmias in patients with idiopathic dilated cardiomyopathy (IDC). We examined this property in 15 patients with IDC undergoing cardiac transplantation and in 14 control subjects. An 8 x 8 array with electrodes 2 mm apart was used to determine the electrical activation sequence over a small region of the left ventricular surface. Tissue from the area beneath the electrode array was examined in the patients with IDC. The patients with IDC could be divided into three groups. Group I (n = 7) had activation patterns and estimates of longitudinal (theta L = 0.84 +/- 0.09 m/s) and transverse (theta T = 0.23 +/- 0.05 m/s) conduction velocities that were no different from controls (theta L = 0.80 +/- 0.08 m/s, theta T = 0.23 +/- 0.03 m/s). Group II (n = 4) had fractionated electrograms and disturbed transverse conduction with normal longitudinal activation, features characteristic of nonuniform anisotropic properties. Two of the control patients also had this pattern. Group III (n = 4) had fractionated potentials and severely disturbed transverse and longitudinal propagation. The amount of myocardial fibrosis correlated with the severity of abnormal propagation. We conclude that (a) severe contractile dysfunction is not necessarily accompanied by changes in propagation, and (b) nonuniform anisotropic propagation is present in a large proportion of patients with IDC and could underlie ventricular arrhythmias in this disorder.

Estimates of repolarization dispersion from electrocardiographic measurements.

BACKGROUND: Repolarization dispersion (Rd) is frequently mentioned as a predictor of cardiac abnormalities. We present a new measure of Rd based on the root-mean-square (RMS) curve of an ECG lead set and compare its performance with that of the commonly used QT dispersion (QTd) measure with the use of recovery times measured from directly recorded canine electrograms. METHODS AND RESULTS: Using isolated, perfused canine hearts suspended in a torso-shaped electrolytic tank, we simultaneously recorded electrograms from 64 epicardial sites and ECGs from 192 "body surface" sites. RMS curves were derived from 4 lead sets: epicardial, body surface, precordial, and a 6-lead optimal set. Repolarization was altered by changing cycle length, temperature, and activation sequence. Rd, calculated directly from recovery times of the 64 epicardial potentials, was then compared with the width of the T wave of the RMS curve and with QTd for each of these 4 lead sets. The correlation between T-wave width and Rd for each lead set, respectively, was epicardium, 0.91; body surface, 0.84; precordial, 0.72; and optimal leads, 0.81. The correlation between QTd and Rd for each lead set was epicardium, 0.46; body surface, 0.47; precordial, 0.17; and optimal leads, 0.11. CONCLUSIONS: RMS curve analysis provides an accurate method of estimating Rd from the body surface. In contrast, QTd analysis provides a poor estimate of Rd.

Variability of the body surface distributions of QRS, ST-T and QRST deflection areas with varied activation sequence in dogs.

Distributions of QRS, ST-T and QRST areas of 192 lead body surface ECG's were measured in dogs for multiple activation orders. Qualitatively, the distributions of QRST area were found to be strikingly similar over all activation orders in contrast to the distributions of QRS or ST-T areas. Quantitative results showed that variability of the QRST areas over all activation orders was consistently less than those of either QRS or ST-T. The factor responsible for the QRS deflection is ventricular activation sequence while those responsible for the ST-T deflection are both activation sequence and ventricular recovery properties. Since the total QRST deflection area was largely independent of activation sequence it is likely the quantity is an index of ventricular recovery properties. The significance of this relation is that QRST deflection area may permit evaluation of intrinsic ventricular recovery properties in the presence of abnormal ventricular activation as occurs with intraventricular conduction disorders and ectopic origin of excitation. Evaluation of intrinsic ventricular recovery properties may also permit recognition of states at risk of ventricular arrhythmias due to increased disparity of these properties.

Diagnosis of old inferior myocardial infarction by body surface isopotential mapping.

Body surface isopotential maps obtained from 28 patients with old inferior wall myocardial infarction were compared with maps from 120 normal subjects. The 12 lead electrocardiogram of 8 of the 28 patients (29 percent) with inferior wall infarction was normal or showed only nondiagnostic ST-T wave abnormalities at the time the isopotential maps were obtained. In all patients with inferior wall infarction the isopotential map showed a minimum (area of negative potentials) on the inferior or right thoracic surface during the early portions of the QRS complex. This finding was observed in patients with normal or nonspecific abnormalities in the 12 lead electrocardiogram as well as those with QRS abnormalities. By contrast, the minimum during the early QRS complex in normal subjects was located on the right upper back and shoulder region...

Application of body surface mapping to exercise testing: S-T80 isoarea maps in patients with coronary artery disease.

Body surface electrocardiographic maps were recorded before and after exercise in 25 men with angiographically documented coronary disease. Torso potential distributions at 192 locations were derived from a 32 lead electrode array using methods previously described in our laboratory. The S-T segment was characterized by the spatial distribution of the integral of S-T segment voltage over 80 ms (S-T80). Body surface regions where the S-T80 areas were =8 mV . ms or greater were identified in 18 of 25 patients. The most negative S-T80 site on the map was called the "S-T80 minimum." The S-T80 minima were located 1 or 2 electrode rows away from the standard V4--V6 electrode positions in 6 of 18 patients who developed S-T80 areas of -8 mV . ms or greater. Our data suggest that standard electrocardiographic leads may not be optimal for identifying S-T segment depression in all patients with coronary disease. Furthermore, body surface mapping during exercise provides a more quantitative and qualitative method for characterizing the ischemic response to exercise.

Electrocardiographic body surface potential mapping.

This article is intended as a review of the field of electrocardiographic body surface potential mapping. The problems associated with measurement of potential distributions will be described, along with the various techniques of data acquisition and display that have evolved over the past decade. Description of mapping applications as they apply to clinical and experimental electrocardiography are presented. The article has been written from a historical perspective, although emphasis is placed on current, state-of-the-art approaches.

Nonparametric identification of discriminative information in body surface maps.

A nonparametric method, based on the Kolmogorov-Smirnov (K-S) test was used to detect significant differences between classes of body surface maps (BSPM's). By systematic application of the method throughout the cardiac cycle, discriminative spatio-temporal information can be identified. In a second method, a Sebestyen linear transformation (SLT) was derived to give estimates of pairwise, linear separability of clinical classes. The utility of the method was illustrated by the pairwise comparison of 40 normal subjects (NOR), 40 patients with anterior myocardial infarction (AMI), and 40 with inferior myocardial infarction (IMI). The application examples demonstrated that: a) diagnostic information in low potential amplitude regions may surpass that in high amplitude regions, b) probability distributions of characteristic features showed small overlap in NOR versus AMI and NOR versus IMI dichotomies although they were not linearly separable, and c) the single best separating potential sample in the K-S sense for NOR versus AMI or NOR versus IMI dichotomies recovered 88 and 73% of the SLT performance, respectively.

Criteria for local myocardial electrical activation: effects of electrogram characteristics.

Detection of local electrical myocardial activation by means of extracellular recordings is often difficult in the presence of polyphasic electrograms. The purpose of this investigation was to compare the ability of several variables to distinguish unipolar deflections due to local activation from those due to nonlocal activity. A model of polyphasic deflections based on atrial recordings during reentrant tachycardia was used to facilitate distinction of local and distant activity by methods independent of the test variables. The performance of variables were assessed by comparing areas under receiver operating characteristic curves. Optimal thresholds of test variables were identified by maximizing statistics which corrected for the pretest probability of local activation. We found that the greatest negative first derivative of the unipolar potential discriminated between local and distant ventricular signals, but performed less well than the ratio of the first derivative to the potential for distinguishing between local atrial signals and distant ventricular signals. A linear combination of the potential and the ratio of the first derivative and the potential performed well for all groups of signals studied. We conclude that optimal criteria for detecting local activation depends on the characteristics of the population of signals and that a statistical approach can be used to identify optimal criteria for a given population.

Anatomical architecture and electrical activity of the heart.

In most early studies of cardiac electrophysiology, the correlation between propagation of excitation and the architecture of cardiac fibers was not addressed. More recently, it has become apparent that the spread of excitation, the sequence of recovery, the associated time-varying potential distributions and the intra- and extracardiac electrocardiograms are strongly affected by the complex orientation of myocardial fibers. This article is a review of older and very recent, partly unpublished, mathematical simulations and experimental findings that document the relationships between cardiac electrophysiology and fiber structure. Important anatomical factors that affect propagation and recovery are: the elongated shape of myocardial fibers which is the basis for electrical anisotropy; the epi-endocardial rotation of fiber direction in the ventricular walls; the epi-endocardial obliqueness of the fibers ("imbrication angle"), and the conduction system. Due to the complex architecture of the fibers, many different pathways are available to an excitation wavefront as it spreads from a pacing site: the straight line; the multiple, bent pathways resulting from the epi-endocardial rotation of fiber direction; the coiling intramural pathways associated with the "imbrication" angles (Streeter) and the pathways involving the Purkinje network. Only in a few cases is the straight line the fastest pathway. The shape of an excitation wavefront at a given time instant results from the competition between all possible pathways. To compute the potential distributions and ECG waveforms generated by a spreading excitation wave we must know the successive shapes and positions of the wavefront, the architecture of the fibers through which it propagates and the spatial distribution of their anisotropic electrical properties.

[Changes in the ECG and myocardial blood flow of the right and left ventricular walls under acute pulmonary arterial stenosis].

For studying the effect of regional myocardial blood flow changes on the epicardial ECG of right and left ventricular walls under acute right ventricular pressure overload, we mapped the epicardial using 64-channel shock electrodes, and estimated the myocardial blood flow with radioactive microspheres. In 9 anesthetized open-chest dogs, the main pulmonary artery was gradedly constricted to the level of mild (peak RV pressure: PRVP, 50-70 mmHg), moderate (PRVP, 70-80 mmHg) and severe stenosis (PRVP, over 80 mmHg). Labeled microspheres were injected into the left atrium before and after the PA constriction, and the epicardial ECGs were recorded continuously. After the completion of the experiment, 9 areas of each right and left ventricular wall were excised. The myocardium was divided into three layers and the flow data were compared to the changes of ECG parameters. In the cases where there was severe PA stenosis, the right ventricular myocardial blood flow decreased to a significantly greater degree (63% reduction from the control), especially in the subepicardial layer, than the flow in the left ventricle (37% reduction from the control). ST potential, STT and QRST Area Map increased in the right ventricle but decreased in the left ventricle. Activation Recovery Time of the right ventricle decreased due to the severe ischemia of the right ventricle. The value of QRS Area Map of the left ventricle decreased significantly in parallel with the decrease in cardiac output.(ABSTRACT TRUNCATED AT 250 WORDS)

Uncertainty of the electrocardiogram: old and new ideas for assessment and interpretation.

The electrocardiogram (ECG) is a highly complex, dynamic and stochastic phenomenon. Although it provides a valuable, noninvasive and rapid means of assessing cardiac state and its change, uncertainties in its measurement and variation in the underlying electrophysiology that generates the ECG make difficult further improvement in its reliability for detecting and monitoring cardiac pathologies and conditions. This article reviews the sources of variability and uncertainty in ECG measurement and interpretation, revisits some old ideas for dealing with them, and proposes some novel directions for improving accuracy of ECG assessment and interpretation. We shall explore relative information content of lead systems, representation of ECG signals and patterns, and estimation of ECG distributions from limited lead systems. In addition, we will compare strategies for measuring ECG information and suggest new paradigms for feature extraction that reduce the sensitivity of assessment accuracy to intrinsic and extrinsic measurement errors. Finally, we review the importance of including dynamic information in ECG assessment, both for interpreting current cardiac state as well as for monitoring its change and significance.

Karhunen-Loeve representation of ECG data.

Quantitative representation of mathematical functions or random data provides a useful and often necessary step for analyzing complex physical or physiological phenomena. In general, the representation process converts complex functions or observations into weighted combinations of simple, elementary components. Rationale for use of representation lies in the simplification of restructuring the original functions or data in terms of these basic elements, which can facilitate understanding or analysis of the phenomena under study. Representation provides a basis for feature extraction in classification problems, filtering for noise reduction, and information assessment. In this paper the author describes some of the classical representation methods and demonstrates specific implementation of the statistical representation of electrocardiographic data using the Karhunen-Loeve method.

Mechanisms in T-wave alternans caused by intraventricular block.

It is recognized that 2:1 intraventricular (IV) block can result in T-wave alternans but is usually assumed that it would also affect QRS waveform. Block in a local region is not, however, varied activation sequence of the same muscle mass because the blocked region is not activated and is not part of the mass that is activated in cycles without block. Also, the block region may have electrocardiogram (ECG) effects when its state differs from other regions. In view of those considerations, the ECG effects of IV block were evaluated by using a computer model of excitation and recovery. ECGs were calculated from differences between the excited state and various degrees of recovery. Results provided evidence that boundaries associated with regions of block rather than regions having varied activation sequence were the major factors in T-wave alternans caused by IV block. Effects of the boundaries included cancellation of the effects of IV block on QRS complexes. Findings suggest that IV block cannot be excluded as a mechanism of T-wave alternans in the absence of QRS alternans.

Entrainment of reentrant tachycardia in a computer model.

Capture of the cardiac rate by pacing followed by an immediate return to the original rate after pacing has been proposed as characteristic of reentrant rhythms. In this study, such entrainment has been demonstrated using computer-model simulations of propagated excitation and of reentry associated with structural and functional obstacles. With structural obstacles, the mechanism of entrainment was bidirectional propagation of paced excitation in reentry circuits, with collision of the reentrant and paced excitation in one direction and continued propagation of paced excitation in the other direction. The time of pacing onset, rate, and location all affected the QRS waveform during entrainment. With a particular time of onset and rate of pacing, the duration of time during which the QRS waveform underwent dynamic change was directly related to the distance between the pacing site and reentrant circuit. The location of reentry associated with functional obstacles moved so that the relationship between pacing-induced and reentrant excitation varied. In some cycles, pacing did not alter reentrant circuits, that is, entrainment did not occur, while other cycles were entrained, but by a different mechanism than that with structural obstacles. Leading circle reentry circuits, consisting of propagation away from and returning to reentry sites, did not have an excitable gap and paced excitation did not enter those circuits. Paced excitation did, however, enter the propagation paths between leading circle reentry circuits and modified the circuits by affecting the recovery of excitability.

Cycle-length effects on the initiation of simulated torsade de pointes.

The sequence of a short and long cycle length has been observed frequently to precede the onset of torsade de pointes in patients. The purpose of this study was to determine the mechanism of that relation by means of a computer model of propagated excitation. The model included cycle length-dependent refractoriness and slow propagation during incomplete recovery and has been used previously to document that changing QRS waveform and limited duration of torsade de pointes episodes can be explained by moving sites of reentrant excitation. In this study, a short-long cycle-length sequence due to a ventricular premature response and compensatory pause was shown to prolong the period during which simulated torsade de pointes could be initiated. A premature response that failed to initiate the arrhythmia in the absence of that sequence did so after the sequence. Both the premature ventricular response and compensatory pause of the short-long cycle-length sequence contributed to prolongation of the torsade period, but by different mechanisms. The compensatory pause prolonged refractory periods and increased their disparity by a direct effect of cycle length. The ventricular premature response had similar effects, but these were indirect and due to the activation sequence of the response. When such a response was followed by a supraventricular response, the ventricular cycle length included atrioventricular conduction time and was longer than that with a series of responses of either supraventricular or ventricular origin. In addition to elucidating the mechanism of the short-long cycle sequence relation to torsade de pointes, the findings suggest that long cycles of whatever nature or the sequence of ventricular and supraventricular responses of whatever cycle length may facilitate initiation of the arrhythmia.

Distribution of QRST deflection areas in relation to repolarization and arrhythmias.

QRST area maps were calculated from a computer model of propagated excitation with nonuniform cycle length-dependent recovery. Vulnerability was independently assessed as fibrillation threshold (FT). Separate effects of varied range and mean recovery durations on FT and QRST maps were determined. FT was inversely related to the range of recovery durations and that range was related to QRST area map features including magnitude, nonuniformity, and gradients. Mean recovery duration was directly related to FT but did not alter QRST maps unless changes of recovery duration were localized. Locally decreased mean duration resulted in decreased FT and increased magnitude and nonuniformity and gradients in QRST maps. Locally increased mean duration had similar effects on QRST maps, but FT was increased. Results support the validity of QRST map features as markers of vulnerability due to disparity of recovery duration or locally decreased mean recovery duration but not due to widespread changes of mean duration or locally increased duration.