BODY SURFACE AND EPICARDIAL ECG MAPPING: STATE OF THE ART AND FUTURE PERSPECTIVES.

Bruno Taccardi , Bonnie B Punske
Nora Eccles Harrison Cardiovascular Research and Training Institute (CVRTI), University of Utah , 95 S. 2000 E.
Salt Lake City, UT 84112,  Utah USA

Abstract: Body surface maps, when obtained from electrocardiograms recorded from the entire surface of the torso, with a sufficient number of leads, contain the entire electrical information that can be captured from body surface measurements. Thousands of studies have shown that Body Surface Maps (BSMs), in their different formats, have higher diagnostic power than the 12-lead ECG. However, BSMs are difficult to interpret in terms of intracardiac electrical events (time course of excitation and recovery and potential distributions) because they don’t provide a direct view of the heart with superimposed excitation and recovery isochrones and potential distributions. Inverse procedures do provide such an image. Starting from body surface electrocardiograms, inverse procedures reconstruct and display epi- and endocardial potentials and isochrones, and show promise to become a new imaging method of great scientific and practical utility, particularly when the interpretation of the images is supported by knowledge of fiber architecture and by accurate mathematical simulations.

Introduction

The imaging methods currently available to cardiologists visualize most aspects of cardiac anatomy and function: the shape, position and size of the heart, the motion of the cardiac walls, the coronary circulation, the regional blood supply, the blood flow in the cavities and great vessels and the spatial distribution of biochemical processes.

Unfortunately, we don't have a clinically practicable method for visualizing the electrical activity of the heart, namely, the spread of excitation, the sequence of recovery and the associated potential and current distributions in the heart and torso. The classical 12-lead electrocardiogram (ECG) provides only a fraction of the total information. Body surface maps (BSMs) contain all the electrical information that can be elicited from body surface measurements, but do not directly visualize the sequence of excitation and recovery in the heart. This is a significant gap in our diagnostic equipment, because the sequence of excitation governs the time course of mechanical contraction and therefore the pumping function of the heart. Knowing and visualizing the sequence of repolarization is also important because local or diffuse disturbances of recovery are often associated with increased vulnerability to cardiac arrhythmias [1]

The heart vector

Early attempts to define the cardiac electric sources from body surface potential maps suggested that the cardiac electrical generator, as seen from the body surface, was equivalent to a single current dipole located in the chest [2]. The dipole could be mathematically represented by a vector (the heart vector), whose magnitude, orientation and polarity could be derived from as few as two ECGs (frontal vector) [3] or three ECGs (spatial vector) [4]. From these studies, many cardiologists drew the conclusion that the three leads that define the spatial vector contained all the information that could be extracted from body surface measurements. Therefore, additional ECG leads were unnecessary and redundant.

Body surface maps

However, in the 1950s and 1960s, more advanced mapping techniques revealed that the potential distributions generated by human and animal hearts were much more complex than those produced by a single dipole [5,6]. In both normal subjects and cardiac patients, BSMs showed multiple features, such as potential maxima and minima, ridges, saddles and valleys that varied in shape, amplitude and location during excitation and recovery. These features changed in characteristic ways in most heart diseases [7]. These findings suggested that important diagnostic information might be obtainable from BSMs. During the last 50 years, specialized centers for the clinical and experimental study of BSMs were created in Asia, Europe, Oceania and the Americas. In addition to potential maps, other types of maps were developed: integral maps, gradient maps, laplacian maps, pseudo-isochrone maps, ARI (activation-recovery interval) maps etc. [1, 7, 8].  Maps were recorded in thousands of patients with congenital and acquired heart diseases: coronary heart disease, conduction defects, atrial and ventricular enlargement, cardiomyopathies, reentrant and focal arrhythmias, long QT syndrome etc. Maps were interpreted using visual inspection, pattern recognition, statistical analysis (significant departure from normal patterns), principal component analysis and pace mapping [7 - 10].

In the great majority of heart conditions, BSMs demonstrated a higher diagnostic power than the 12-lead ECG. In many cases, significant features, such as abnormal potential minima or maxima, signaling myocardial ischemia or infarction, were found in chest areas that are not sampled by the 12-lead system. In those cases, the classical 12-lead ECG was normal and overlooked the pathological features (See Ref [7], Page 1027-1031)..

Limitations of body surface maps

Despite their demonstrated diagnostic power, BSMs have not become a routine clinical method. This is due to a number of difficulties: necessity of recording a high number of leads (32 to 219); multiplicity of lead systems used by different laboratories; difficulty of memorizing hundreds of patterns (usually one for every msec during the P-QRST interval) that vary in different heart diseases. The most serious obstacle, however, is the difficulty of interpreting the various features of a map (number, location, amplitude and time-course of maxima, minima, saddles etc.) in terms of intracardiac events. Most of the other imaging methods provide a direct representation of cardiac anatomy and function that is easily recognizable by cardiologists. Conversely, BSMs do not show an image of the heart, but an attenuated, smoothed and distorted reflection of the intracardiac electrical events, as projected on the body surface. Thus, many important features of the electrical activity of the heart (multiple epicardial breakthrough sites, shape and density of excitation and recovery isochrones, reentry circuits in tachycardias, effects of anisotropy on wave front propagation and potential patterns) are not recognizable through visual inspection of the maps.

How to overcome the limitations of Body Surface Maps

The best way to obtain images that accurately represent the cardiac electrical activity consists of recording electrograms (EGs) directly from the heart, using dense arrays of epicardial, intramural or intracavitary electrodes. From the recorded signals we can construct three-dimensional images that represent the heart with superimposed equipotential lines, isochrone lines etc. [11].  These maps are much easier to interpret than body surface maps because they display the anatomical shape of the heart together with the propagating excitation waves and the time-varying potential distributions. Obviously, this approach requires invasive procedures, and can be implemented only at surgery or in experimental preparations.

A noninvasive approach consists of solving the “inverse problem of electrocardiography”, that is, mathematically reconstructing the distribution of electrical events in the heart from body surface ECGs. A full reconstruction of the intracardiac current sources is not possible because the problem has no unique solution [12].  However, if we limit the scope of the inverse procedure to reconstructing the potential distribution or the sequence of excitation on the epicardial and/or endocardial surface, the problem does have a unique solution. Two main methods have been followed: a) direct reconstruction of epicardial and endocardial isochrones [13]. This method is based on the “uniform dipole layer” model of the excitation wave front, and on the “cardiac surface” equivalent generator [14]. Validation of this procedure has been performed in human subjects by comparison with measured isochrones [15] and in model studies [16]. The method requires knowledge of the epicardial and endocardial geometry.  In case of myocardial infarction, the geometry of the infarcted area  must be known as well.

Method b) consists of computing epicardial potentials from body surface electrocardiograms. This is achieved by finding a transfer matrix such that the potential at each site on the epicardium can be computed as a linear combination of weighted potentials from multiple body surface sites. Due to the ill-posed nature of the inverse problem, the solution must be constrained and smoothed by means of regularizing or optimizing procedures [12, 17]. Many studies, pursued in cooperation with Dr. Yoram Rudy’s group in Cleveland, have shown the feasibility of reconstructing epicardial potentials, isochrones and electrograms (EGs) in torso-shaped electrolytic tanks containing an isolated heart. Good reconstructions were obtained in normal hearts and hearts with acute or 5-day old myocardial infarction, during atrial and ventricular drive and during ventricular tachycardias [18]. The computed maps accurately revealed the location and intramural depth of a pacing site, the size and location of an infarcted area, the presence of electrical activity in the surviving “border zone”, the epicardial portion of the reentry circuit in reentrant tachycardias and the distribution of propagation velocities on the epicardium. Validation of the method was performed through comparison with measured epicardial data [18]. The same approach has been used in a limited number of human subjects, with promising results [19, 20]. Endocardial potentials, isochrones and EGs were also reconstructed from non-contact multi-electrodes intracavitary probes in experimental animals [21] and human subjects [22].

Epicardial and endocardial mapping

The inverse procedures show promise of offering the cardiologist a three-dimensional image of the patient’s heart with superimposed epicardial or endocardial isochrones, time-varying potential distributions, reentry circuits during tachycardias and an outline of the infarcted areas, on a beat-by-beat basis. The next step must consist of interpreting these surface images in terms of intramural activity. As previously mentioned, a complete reconstruction of the intramural activity from extracardiac measurements cannot be expected from the currently available inverse procedures, because the problem has no unique solution. However, studies from our and other groups have shown that by mapping epicardial and endocardial potentials and isochrones, with high spatial resolution, during sinus rhythm and ventricular pacing, a great deal of information can be obtained, that provides some insight into deep intramural events. Interpretation of epicardial and endocardial maps is greatly facilitated if we know the architecture of myocardial fibers and their anisotropic electrical properties. Previous work showed that the epicardial potential and isochrone patterns [23, 24], combined with knowledge of the anisotropic properties and intramural rotation of the fibers, makes it possible to assess the location and depth of an intramural pacing site, the direction of myocardial fibers near the pacing site, the amount of intramural rotation of the fibers, the presence of small, non-transmural necroses, and the effects of the “imbrication” angle [25] on the spread of excitation. Moreover, the distribution of propagation velocities on the epicardium during atrial and ventricular stimulation show characteristic patterns that can be interpreted in terms of deep activity if we know the rules that govern the three-dimensional propagation of excitation, in particular the effect of fiber direction, shape and curvature of the wave fronts, collision with boundaries and involvement of the conduction system on the spread of excitation [23]. Considerable insight into the relationships between epicardial (or endocardial) and deep, intramural activity has been obtained by performing 3-D explorations of the ventricular myocardium by means of multiple intramural needles, combined with histological analysis of fiber architecture [26]. More detailed representations of the intramural, epicardial and endocardial propagation and associated potential fields have been obtained by means of numerical simulations that reproduce the shape of the ventricles, the architecture of the fibers and their anisotropic properties [27]. Comparison of simulations with experimentally measured data showed that the most accurate representations of propagation and associated potentials are based on the anisotropic bidomain model with unequal anisotropy ratio.

Conclusion

Body surface maps, when recorded from the entire surface of the torso with a sufficient number of leads, contain the entire electrical information that can be captured from body surface measurements. They have higher diagnostic power than the 12-lead ECG. However, BSMs are difficult to interpret in terms of intracardiac electrical events (time course of excitation, recovery and potential distributions) because they do not provide a direct view of the heart with superimposed isochrones and potential distributions. Inverse procedures, which reconstruct and depict epi- and endocardial potentials and isochrones, show promise to become a new imaging method of great scientific and practical utility, particularly when the interpretation of the images is supported by knowledge of fiber architecture and by accurate mathematical simulations.

ACKNOWLEDGEMENT

This work was supported by NIH Grant RO1 HL42376 and by awards from the Nora Eccles Treadwell Foundation and the Richard A. and Nora Eccles Harrison Fund for Cardiovascular Research.

REFERENCES

[1] Takada Y, Akahoshi M, Inden  Y, Shibata Y, Shimitzu A, Yoshida Y, Yamada T, Tsuboi N, Hirayama H, Ito T,  Kondo T, Saito H, Hirai M: Changes in repolarization properties with long term cardiac memory modify dispersion of repolarization in patients with Wolff-Parkinson-White syndrome. J Cardiovasc Electrophysiol April 2002.

[2] Waller AD. On the electromotive changes connected with the beat of the mammalian heart and of the human heart in particular. Philos Trans R Soc. 1889;B 180:169-194.

[3] Einthoven W, Fahr G, de Waart A. Über die Richtung und manifeste Grösse der Potentialschwankungen im menschlichen Herzens und über den Einfluss der Herzlage auf die form des Elektrocardiogramms. Pflügers Arch ges Physiol. 1913;150:275-315.

[4] Wilson FN, Johnston FD, Kossmann CE: The substitution of a tetrahedron for the Einthoven triangle. Am Heart J 1947;33:594-603

[5] Nahum LH, Mauro A, Chernoff HM, Sikand RS. Instantaneous equipotential distribution on surface of the human body for various instants in the cardiac cycle. J. Appl. Physiol. 1951;3:454-464.

[6] Taccardi B. Distribution of heart potentials on the thoracic surface of normal human subjects. Circ. Res. 1963;12:341-352.

[7] De Ambroggi L, Musso E, Taccardi B. Body surface mapping. In: Macfarlane P, Veitch Lawrie T, eds. Comprehensive electrocardiology. New York: Pergamon Press; 1989:1015-1049.

[8] Flowers N, Horan L. Body surface potential mapping. In: Zipes D, Jalife J, eds. Cardiac electrophysiology. From cell to bedside. Philadelphia: Saunders; 1995: 1049-1067 and 2000:737-746.

[9] Liebman J, Plonsey R, Rudy Y, eds. Pediatric and fundamental electrocardiography. Boston: Martinus Nijhoff; 1986.

[10] van Dessel PF, van Hemel NM, de Bakker JM, Linnenbank AC, Potse M, Jessurun ER, SippensGroenewegen A, Wever EF. Relation between body surface mapping and endocardial spread of ventricular activation in postinfarction hearts. J Cardiovasc Electrophysiol. 2001;12:1232-1241.

[11] Arisi G, Macchi E, Corradi C, et al: Epicardial excitation during ventricular pacing.  Relative independence of breakthrough sites from excitation sequence in canine right ventricle. Circ Res 1992; 71(4):840-849.

[12] Gulrajani RM: The forward and inverse problems of electrocardiography. IEEE Eng Med Biol Mag. 1998;17:84-101,122

[13] Huiskamp G, Van Oosterom A. The depolarization sequence of the human heart surface computed from measured body surface potentials. IEEE Trans Biomed Eng. 1988;35:1047-58

[14] Simms HD, Jr., Geselowitz DB: Computation of heart surface potentials using the surface source model. J Cardiovasc Electrophysiol 1995; 6(7):52-531.

[15] Oostendorp T, Pesola K. Non-invasive determination of the activation sequence of the heart: validation by comparison with invasive human data. In: Computers in Cardiology. Cleveland, Ohio: IEEE Computer Society Press; 1998:313-316.

[16] Pullan AJ, Cheng LK, Nash MP, Bradley CP, Paterson DJ. Noninvasive electrical imaging of the heart: theory and model development. Ann Biomed Eng. 2001;29:817-836.

[17] Ramanathan C, Jia P, Ghanem RN, Calvetti D, Rudy Y. Noninvasive electrocardiographic imaging (ECGI) using the generalized minimal residual method (GMRES). Ann. Biomed. Eng. 2001;29 (Suppl. 1):S-47 (Abstract).

[18] Burnes JE, Taccardi B, Ershler PR, Rudy Y. Noninvasive ECG imaging of substrate and intramural ventricular tachycardia in infarcted hearts. J Am Coll Cardiol. 2001;38:2071-2078.

[19] Shahidi AV, Savard P, Nadeau R. Forward and inverse problems of electrocardiography: modeling and recovery of epicardial potentials in humans. IEEE Trans Biomed Eng. 1994;41:249-56.

[20] MacLeod RS, Gardner M, Miller RM, Horacek BM. Application of an electrocardiographic inverse solution to localize ischemia during coronary angioplasty. J Cardiovasc Electrophysiol. 1995;6:2-18.

[21] Khoury DS, Taccardi B, Lux RL, Ershler PR, Rudy Y. Reconstruction of endocardial potentials from measured intracavitary potentials for paced rhythms in the isolated canine left ventricle. Circulation. 1995;91:845-863.

[22] Paul T, Windhagen-Mahnert B, Kriebel T, Bertram H, Kaulitz R, Korte T, Niehaus M, Tebbenjohanns J.. Atrial reentrant tachycardia after surgery for congenital heart disease: endocardial mapping and radiofrequency catheter ablation using a novel, noncontact mapping system. Circulation 2001;103:2266-2271

[23] Taccardi B, Lux R, Ershler P, et al: Potential distributions and excitation time maps recorded with high spatial resolution from the entire ventricular surface of exposed dog hearts. Computers in Cardiology 1992, Durham, NC, 1992.

[24] Taccardi B, Macchi E, Lux RL, Ershler PR, Spaggiari S, Baruffi S, Vyhmeister Y. Effect of myocardial fiber direction on epicardial potentials. Circulation. 1994;90:3076-90.

[25] Streeter D.: Gross morphology and fiber geometry of the heart, in: Handbook of Physiology, Vol. 1: The Heart, Sect. 2: The Cardiovascular System. Berne R.M. (Ed.), Chapt. 4, 61-112. Baltimore, Williams and Wilkins, 1979

[26] Taccardi B, Lux RL, Ershler PR, et al: Anatomical architecture and electrical activity of the heart. Acta Cardiologica 1997; 52(2):91-105.

[27] Colli Franzone P, Guerri L, Pennacchio M, et al: Spread of excitation in 3-D models of the anisotropic cardiac tissue. Part II. Effects of geometry and fiber architecture of the ventricular wall. Math Biosc 1998; 147:131-171.