| Grant number: | 13/18158-0 |
| Support Opportunities: | Scholarships in Brazil - Doctorate |
| Start date: | November 01, 2014 |
| End date: | April 30, 2017 |
| Field of knowledge: | Health Sciences - Collective Health - Epidemiology |
| Agreement: | Coordination of Improvement of Higher Education Personnel (CAPES) |
| Principal Investigator: | Aluisio Augusto Cotrim Segurado |
| Grantee: | Alex Jones Flores Cassenote |
| Host Institution: | Faculdade de Medicina (FM). Universidade de São Paulo (USP). São Paulo , SP, Brazil |
Abstract Introduction: the advent of highly active antiretroviral therapy was observed profound impact on the natural history of HIV infection. Thus, the use of combination therapies containing different classes of drugs promoted a significant and sustained suppression of viral replication, increasing the survival and quality of life in AIDS patients. Admittedly this therapy has led many of the observed changes in the pattern of disease by introducing new phenomena, so the continuous surveillance of emerging standards is so vital. The pathophysiologic mechanism of metabolic changes in HIV/AIDS patients is still not fully understood and therefore consensus specific to your treatment are not yet available for proper orientation of health professionals. Objectives: this proposal aims to analyze and characterize the endocrine and metabolic changes associated with the use of antiretroviral drugs in people living with HIV/AIDS and developing intelligent computational algorithms aiming its identification and prediction. Methods: data from a retrospective cohort study to assess the heath damage from the use of antiretroviral drugs for people living with HIV/AIDS will be used. Kaplan Mayer and Cox regression will be used to characterize the time until the occurrence of metabolic and endocrine disorders, as well as the impact of the factors associated with these disorders. Linguistic model using fuzzy logic is built by means of computational rules using operators such as "AND" e "OR" to connect the antecedent situation with the consequent respose. The paraconsistent artificial neural network (PANN) will be developed in accordance with the pattern of occurrence of metabolic and endocrine disorders and associated factors; it will be structured in layers connected to each other for better management of information. The algorithms are constructed with portion of 50% of subjects in the database and the remaining records are used to the respective tests. The tests are performed on the accuracy of the algorithms compared with the results observed in the original database. Expected results: we hope that the intelligent algorithms applied to the identification and prediction of endocrine and metabolic changes may serve as a support tool to help health professionals to minimize the uncertainties about these changes, maximizing their power of decision-making based on statistical evidence and the expertise from specialists. (AU) | |
| News published in Agência FAPESP Newsletter about the scholarship: | |
| More itemsLess items | |
| TITULO | |
| Articles published in other media outlets ( ): | |
| More itemsLess items | |
| VEICULO: TITULO (DATA) | |
| VEICULO: TITULO (DATA) | |