Full text
| |
| Author(s): |
Neli Regina de Siqueira Ortega
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | São Paulo. |
| Institution: | Universidade de São Paulo (USP). Instituto de Física (IF/SBI) |
| Defense date: | 2001-05-24 |
| Examining board members: |
Eduardo Massad;
Nestor Felipe Caticha Alfonso;
Rodney Carlos Bassanezi;
Fernando Antonio Campos Gomide;
Dirce Maria Trevisan Zanetta
|
| Advisor: | Eduardo Massad |
| Abstract | |
Biological, medical and epidemic systems present several types of inherent uncertainties to its processes. Many of these uncertainties have been treated in an efficient way with statistical and Bayesian models. Though, these areas still lack of mathematical structures that make possible the treatment of the non-statistical uncertainties typical of some of these systems. Besides, the use of linguistic terms to express quantitatively the variables is very common in these areas. So, due to its features, the fuzzy logic comes as an appropriate theory to treat some of these problems. The aim of this thesis was to develop applications of the fuzzy logic theory to problems of biomedicine. Our challenge was to propose paths, to look for ways, of accomplishing an effective junction off this theory with the mentioned areas, mainly with epidemiology. Eight works were elaborated, where several aspects of this theory were approached, such as: static and dynamic fuzzy linguistic models, fuzzy decision making, probability of fuzzy events, fuzzy relations and the use of extension principle in the construction of fuzzy rules. We conclude that the fuzzy logic theory can aid in the treatment of many epidemiological problems, as well as diagnostic systems. We also show that it can work effectively in decision making processes of Public Health. Some systems can, pottentially, help the physicians in the diagnosis and prognostic of diseases, mainly in the specialists absence. The static linguistic models worked well, however, difficulties concerning the dynamic models still need to be overcome. We also discuss the specialists role in the elaboration of fuzzy models in epidemiology and propose a method for elaboration of less dependent models. All the works presented good results, stimulating the continuity of the researches in this area. (AU) | |