A NEW INDEX FOR THE QUANTITATION OF THE VENTRICULAR REPOLARIZATION

E. V. Garcia¹, J. L. B. Marques¹
¹Instituto de Engenharia Biomédica – CTC, Universidade Federal de Santa Catarina
Campus Universitário Trindade, Florianópolis, SC BRAZIL 88040-900

Abstract: This paper presents a new variable for the quantitative analysis of the ventricular repolarization: the Asymmetry Factor (FA), which is based solely on the morphological aspects of the T wave. The FA is intended to add new insights to this field of electrocardiology, since even nowadays the investigation of the ventricular repolarization is mostly centered in the QT interval analysis.

INTRODUCTION

Although the underlying physiological mechanisms are not well-established yet, it is clear now the relationship between hypoglycaemia and some changes in the ventricular repolarization such as: increase of the QT interval; flattening of the T wave; TU fused waves [1][2]. Nevertheless, the lack of valid clinical parameters hindered the advances towards an ECG-based hypoglycaemia detector. Harris et al. [3], pointed out that just QT interval analysis is not enough for the discrimination between healthy and hypoglycaemic states. His proposed solution was the study of the T wave morphology.

This T wave modeling has been tried by different approaches: pole-zero models [4]; sum of generated waves [5]; and time intervals and calculus of areas [6]. Regarding to ease of comprehension and clinical utility, the last method is the best.

METHODS

We derived a new index for the ventricular repolarization quantitation – the Asymmetry Factor (FA) – from the area under the T wave. Indeed, the FA is based on the ratio between the area posterior to the peak value of the T wave and the total area under the curve [7]. Assigning the name S₂ to this ratio, we can calculate the FA according to (1).

(1)

Unlike other symmetry estimators, the FA not only carries information about the proportion between the areas under each T wave side but also indicates which is the major part. Some examples of the T wave morphology are described in Table I.

TABLE I
Examples of T wave morphology

FA

Morphology

0

Symmetric T wave

-2

Initial T wave area is 1.5 times the posterior area
(S₂ = 40%; S₁ = 60%)

+5

Posterior T wave area is 1.5 times the initial area
(S₂ = 75%; S₁ = 25%)

A preliminary evaluation was done by processing three 5-minute ECG data files. All were recorded from healthy male subjects of 23, 25 and 36 years. The sampling frequency was 250 Hz and the data were digitized in a 12-bit resolution.

RESULTS

TABLE II
Mean values (mean ± SD) of the ratio (posterior area/ total area) and the FA

Subject

S2(%)

FA

(1)

31.16 ± 5.51

-3.77 ± 1.09

(2)

32.68 ± 4.62

-3.46 ± 0.92

(3)

34.28 ± 3.29

-3.14 ± 0.66

Mean

32.71 ± 4.48

-3.45 ± 0.89

TABLE III
FA percentual distribution grouped by classes and sorted by subject.

Classes

Subjects

1

2

3

FA ≥ -1.5

0.51

0.80

0.00

-2.5 ≤ FA ≤ -1.5

11.62

16.06

15.06

-3.5 ≤ FA ≤ -2.5

29.29

35.74

56.37

-4.5 ≤ FA ≤ -3.5

32.83

31.73

25.87

-5.5 ≤ FA ≤ -4.5

18.18

15.26

2.32

-6.5 ≤ FA ≤ -5.5

7.07

0.40

0.00

-7.5 ≤ FA ≤ -6.5

0.51

0.00

0.39

Figure 1: Comparison between Normal Distribution and practical FA probabilities.

DISCUSSION

None of the subjects registered a positive FA, which means the T wave in healthy subjects was always asymmetric, with its initial part predominant over the half posterior to the peak value. Most of the heart beats in any subject had FA values inside the interval -4.5 ≤ FA ≤ -2.5 (Subj.1: 62.12%; Subj.2: 67.47%; Subj.3: 82.24%). In other words, the initial part of the T wave usually was 1.64 to 2.64 times greater than the posterior area.

Figure 1 depicts the plot of the FA mean probabilities and a Gaussian curve calculated with the mean values of this estimator (Table II). Despite the few subjects reported, it still can be noticed a good correlation between the curves.

Further tests are needed to validate this Gaussian model of the Asymmetry Factor, though. They must include not only normal ECG but also data recorded during a hypoglycaemic episode.

Acknowledgments: Work supported by the Brazilian National Council of Scientific and Technological Development.

REFERENCES

[1] J.M.G. Ingelmo, A.N. Angulo, E.H. González, et al.. “Alteraciones electrocardiográficas en la hipoglucemia,” Rev.Esp. Cardiol., vol. 51, pp. 404-406, 1998.

[2] J.L.B. Marques, E. George, S.R. Peacev, et al.. “Altered Ventricular Repolarization During Hypoglycaemia in Patients with Diabetes,” Diabetic Medicine, vol. 14, pp. 648-654, 1997.

[3] R.H. Ireland, R.T.C.E. Robinson, S.R. Heller, et. al.. “QT measurement for TU fused ECG morphology as exhibited during hypoglycaemia,” in Proceedings of the 20^th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 1998, pp. 240-242.

[4] I.S.N. Murthy, G.S.S. Durga Prasad. “Analysis of ECG from Pole-Zero Models,” IEEE Trans. On Biom. Eng., vol. 39, pp. 741-751, 1992.

[5] J.A. Vila, Y. Gang, J.M.R. Presedo, et al.. “A New Approach for TU Complex Characterization,” IEEE Trans. On Biom. Eng., vol. 47, pp. 764-772, 2000.

[6] M. Merri, J. Benhorin, M. Alberti, et al.. “Electrocardiographic Quantitation of Ventricular Repolarization,” Circulation, vol. 80, pp. 1301-1308, 1989.

[7] E.V. Garcia, Instrumentação para a monitoração das alterações eletrocardiográficas decorrentes da hipoglicemia. Florianópolis: UFSC, 2001.