Resumo

RODRIGUEZ, JAVIER et al. PREDICTIONS OF PROPORTIONAL ENTROPY IN CARDIAC DYNAMICS. rev.fac.med [online]. 2015, vol.23, n.2, pp.60-70. ISSN 0121-5256.

Based on dynamic systems theory, a methodology that allows to differentiate normal cardiac dynamics, different levels of abnormality and evolution between these states, was developed. This was done by quantifying the probability of ordered pairs of heart rates in the phase space, through entropy and its proportions. To apply in a clinical setting the developed methodology to 400 cardiac dynamics in order to establish its effectiveness by comparing it with the conventional diagnosis. Holters from 400 individuals were tested for a minimum of 18 hours; 50 with normal diagnosis and 350 with different pathologies. An attractor was built for each one of them in the delay map, and the occupation probability of ordered pairs of heart rates was evaluated through entropy and its proportions. Afterwards, results were compared with values of normality and disease previously established to obtain the diagnosis for each Holter. The findings of Holter and medical history were only revealed after applying the physical-mathematical methodology, in order to calculate sensitivity, specificity and Kappa coefficient regarding to the Gold-Standard. With the entropy proportions of the attractors, acute cardiac dynamics were differentiated from chronic and normal ones, as well as the evolution between normality and disease. A clinical application of the predictive methodology for Holter was developed. Sensitivity and specificity were both 100% and the correlation between the Gold-Standard and the physical-mathematical diagnosis was 1. The application of the methodology allowed establishing quantitatively states of normality and disease of the cardiac dynamic, showing a self-organization of the geometrical dynamic attractor.

Palavras-chave : diagnosis; cardiovascular diseases; adult; heart rate; nonlinear dynamics.

        · resumo em Português | Espanhol     · texto em Espanhol     · Espanhol ( pdf )