Author :
Wang, Qixin ; Liu, Yang ; Mo, Lihong
Abstract :
HIV/AIDS will be one of the greatest challenges in public health in this century. Nowadays the treatment and prevention of AIDS have become one of the most important research areas in the life science. Although there are many ways of therapy available, the therapeutic effects are not definitive. Thus, to achieve individual therapy of AIDS, the evaluation of these therapies becomes necessary. In this passage, using the data of ACTG320 group, we applied grey system theory to construct GM (1, 1) model. Taking into account of both the level of CD4 cells and HIV viral load, we defined K ( i,t ) as the curative effect index and evaluating the treatment. By running the program in Matlab, we estimate the best time frame to withdrawal the drugs for each patient and conclude that for most patients, the best time frame of withdrawal or change to use other drugs is from 20 to 40 weeks after the administration of the drug. Then we analyzed the data of 193A group to evaluate the therapeutic effect of the 4 groups of patients. We divided the patients into two groups: one is the young groups, the other is the old group, and the criterion is 50 year´s old. We did this to reduce the affect of age on the therapeutic effect. In each age group, we used the treatment time and the dosage of drugs of four group patients as training sample input the neural network model to standardize the level of CD4 cells after treatment to evaluate the therapy methods in 193A. To remove the random errors which the neural network model brings into the process of solving, we predict the curative effect of the best therapy of each age group using time series model and run a program in SAS. To better apply our analysis into practice, we modify our goal programming model by Lingo software considering the price of the drugs. We concluded that compare with the therapy of reduce the type of drugs, reducing the dose is the better choice for the young - patients when they are not in a good economic condition.
Keywords :
cellular biophysics; data analysis; diseases; drugs; grey systems; learning (artificial intelligence); mathematical programming; medical computing; neural nets; patient treatment; socio-economic effects; time series; AIDS therapy; CD4 cells; HIV viral load; Lingo software; Matlab; age 50 yr; curative effect index; data analysis; drug administration; drug dosage; economic condition; goal programming model; grey system theory; neural network model; public health; random errors; therapeutic effects; time series model; training sample input; Acquired immune deficiency syndrome; Data analysis; Drugs; Human immunodeficiency virus; Mathematical model; Medical treatment; Neural networks; Predictive models; Public healthcare; Synthetic aperture sonar; AIDS; ARIMA model; neural network model;
