Abstract

RODRIGUEZ, Javier et al. CD4 Lymphocytes Prediction Based on the Theory of Probability: Clinical application on leukocytes, lymphocytes and CD4 populations of HIV patients. Infect. [online]. 2012, vol.16, n.1, pp.15-22. ISSN 0123-9392.

Background: Models for the prediction of the CD4 decrease rate in VIH/AIDS patients have been developed with neural networks, and also for the prediction of the variability in CD4 counts decrease in seropositive and seronegative patients with epidemiological models. Objective: to predict the TCD4 lymphocyte population, beginning with the Complete Blood Count (CBC) information of leucocytes and lymphocytes, through probabilistic theories. Methods: Seven CBC of particular cases were evaluated, by determining the total count of leucocytes and lymphocytes, and the proportion of CD4 Lymphocytes by flow cytometry. Based on a previous work an induction was developed; with that induction the sets A, B, C and D were determined. Based on that sets we established the probability of aparition of specific combinations of leucocytes, Lymphocytes ans CD4 values, in ranges of 1.000 leucocytes, in 128 cases. Results: it is predicted that the leucocytes ranges lower than 5,000 and 4,000 per cubic milimeter of peripheral blood are associated with less than 570 CD4 per microliter with a probability of 0.92 and 1 respectively. As the quantity of populations diminishes, the success percentage between the three measures is more effective. Conclusion: The probability theory revealed a mathematical self-organization of the studied cellular populations, allowing the development of predictions in a clinical level for the CD4 number from the CBC, diminishing costs and deducting the results of the flow cytometry.

Keywords : Probability; VIH; AIDS; CD4; CBC; flow cytometry.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )