Detection of HIV Cellular Evidence by Pattern Recognition Analysis

摘要

More than 80% of AIDS (seropositive) diagnoses during 2001 were among residents of the poorest counties, although these counties represented only one quarter of the population living with diagnosed HIV due to the highly active anti-retroviral therapy (HAART). That is the declines in deaths among persons with AIDS since 2001 were caused primarily by the slower progression of HIV-associated immune deficiency among persons who used HAART. However, AIDS diagnoses and deaths were relatively constant among the poor. The populations in which the HIV epidemic is becoming concentrated (racial, ethnic) need increased access to prevention programs and health care services so the importance of this low cost HIV and AIDS estimator. During the last 10 years, groups with less access to medical care have been affected more and more by the HIV epidemic. To monitoring the status of the epidemic and evaluating the effectiveness of prevention, cheaper HIV diagnoses will be essential. This work presents a novel numerical method of analysis for HIV diagnoses in possible contaminated blood laboratory tests. It turns possible quick estimation in early stages where the numbers of HIV cells are small. The implemented system of pattern recognition is based not only on conventional feature extraction but use also chaotic measurements (fractal dimension and lacunarity estimators). The presented methodology estimates AIDS using secure and relatively low in price techniques. It is very adequate for underdeveloped or developing nations of Africa, Asia, and Latin America where greater the proportion of the population below the poverty level.