FUZZY GENETIC ALGORITHM FOR TELEMEDICINE

Abstract

Uncertainties are always present in the medical diagnosis. In this thesis the uncertainty and vagueness in the Blood Pressure monitoring have been handled by processing them with fuzzy- genetic algorithm. The initial population for Genetic Algorithm applied is generated with assumed fuzzy functions. Genetic algorithm can yield global diagnostic by tuning the fuzzy functions, whatever the weights of the symptoms are considered or fuzzy functions assumed. These algorithms are based on certain concepts of biology and represent the coded set of symptoms, called population. These functions are optimized using theory of reproduction, crossover and mutation. The result shows that fuzzy genetic algorithm shows better results than other methods. A good diagnosis based on the symptoms that a patient display is quite important. The symptom and complaints may be of various degrees and weights. It therefore makes sense to study these problems of diagnosis in the context of fuzzy sets, which can take into account the ambiguity and vagueness. The genetic algorithms are used to optimize the fuzzy functions of global diagnostic.