Recognition of Myocardial Ischemic Lesion on the Basis of ECG-Mapping Data

Leonid I Titomir^a, Vladimir G Trunov ^a, Eduard A I Aidu ^a, Tamara A Sakhnova ^b,
Andrei A Mikhnev ^c

^aInstitute for Information Transmission Problems, Russian Academy of Sciences, Moscow, Russia
^bCardiology Research Complex, Ministry of Health of Russian Federation, Moscow, Russia
^cAll-Russian Research Institute of Automatics, Russian Federation Ministry of Atomic Energy, Moscow, Russia

Correspondence: LI Titomir, Institute for Problems of Information Transmission RAS, B. Karetny 19, 127994 Moscow, Russia.
E-mail: titomir@iitp.ru, phone + 095 209 46 79, fax + 095 209 05 79

1.  Introduction

This work is devoted to one of important problems which are solved in the framework of diagnostic interpretation of the ECG-mapping data, in particular, to recognition of the ischemic lesion, including estimation of its anatomic localization, size, and acuity (degree of the injury development). The proposed method is based on a thin-wall spherical model of the equivalent cardioelectric generator providing efficient application of the multipole expansion of the cardioelectric field and pictorial visualization of the myocardium electrophysiological states [Titomir and Kneppo, 1994]. The initial data are obtained by means of the reduced ECG-mapping lead system NEKTAL-16 with 16 electrodes conveniently situated on the body surface and readily used in practice [Titomir et al., 2001].

2.  Material and Methods

On the basis of well known experimental evidences of the structure of local ischemic lesions, we consider a simplified model of the cardioelectric generator producing the potential distribution with two main opposite extrema on the body surface, as well as a shift of the S-T segment in the electrocardiograms recorded. Here, the secondary shift with respect to the diastolic segment T-Q level is mainly meant, insofar as it is actually displaced from the zero potential level (in principle, this secondary shift may add to the primary systolic shift). It is assumed that the boundary of the uniform double-layer generator surface separating the normal and ischemic myocardium is a circle lying on a sphere with the radius RH, approximating the ventricular wall according to the thin-wall model (Fig. 1). The origin of the coordinate system xyz coincides with the heart center, and its z-axis passes through the center of the circular generator boundary. The imaging sphere (IS) with the radius R is considered to be the surface of a solid homogeneous conducting sphere surrounded by dielectric medium and having the same center. By virtue of the assumed orientation of the generator and symmetry of the model with respect to the z-axis, the potential maximum and minimum on IS are at the intersections of this sphere with the positive and negative z-semiaxes, respectively.

The potential at the IS points with the same polar distance as the points on the generator disk is

(1)

where φ_(0+π)= φ(θ = 0)+φ(θ = π), φ_(0-π)= φ(θ = 0)-φ(θ = π). The points of the ischemic region are projected onto IS at those places where φ exceeds the threshold Δ_φ.

For the potential difference U_D across the equivalent generator surface, which characterizes the degree of injury, the following expression is obtained:

(2)

The maximum generator intensity (moment of the double layer) corresponding to the depolarization front in the normal myocardium, according to the model used is characterized by the potential

(3)

where φ_m is the maximum range of the potential values on IS at the ventricular depolarization period.

3.  Results

The cardioelectric potentials on the body surface were measured with the NEKTAL-16 lead system in a group of patients with verified ischemic lesions localized in various regions of the left ventricle. The potential maps at the middle of the S-T period were considered. Here, the S-T segment shift was observed in some leads, while the body surface potential distributions had obviously “dipolar” shape. To provide the display of the ischemic region on the maps, the potential distribution on IS for were determined, then the threshold potential Δφ and difference potentials UD and UDm were calculated. In Fig. 2, such maps are shown as projected onto IS cut along the right meridian and unrolled in the form of planar isoareal projection. Here, two cases with different localizations of the ischemic lesions are illustrated. The ischemia stages estimated by the ratio UD/ UDm are indicated with different shading.

4.  Discussion and Conclusion

The noninvasive mapping of the cardioelectric potential on a spherical imaging surface and determination of the parameters of hypothetical equivalent cardiogenerator at the middle part of the systolic period make it possible to estimate the main characteristics of the ischemic lesion. These characteristics can be evaluated quantitatively and represented in a pictorial form. An important advantage of the method is the simplicity of the procedure based on using reduced ECG lead system.

References

Titomir LI, Kneppo P. Bioelectric and Biomagnetic Fields. Theory and Applications in Electrocardiology. CRC Press, Boca Raton, 1994.

Titomir LI, Sakhnova TA, Aidu EAI, Trunov VG, Mikhnev AA. Few-lead systems for noninvasive imaging of the cardioelectric field on a standard surface. Biomedizinische Technik, 46(2): 91-93, 2001.