A fuzzy model to predict risk of urinary tract infection

Abstract

Urinary tract infections (UTIs) are among the most common bacterial infections and account for a significant part of the workload in clinical microbiology laboratories. Hence, urine is the specimen most frequently submitted for culture. Physicians distinguish UTIs from other diseases that have similar clinical presentations with use of a small number of tests to distinguish bacteriuria. The microbiological examination of urine consists of examining a methylene smear of the urine specimen, a screening test of significant bacteriuria and culture. In the smear one or more bacterial cells per oil- immersion field usually implies that there are 105 or more bacteria per milliliter in the specimen, the number of RBC and WBC is also a very important indicator. In literature, the normal ranges of these variables are differently defined. The analysis of this data could be very simplified using data management systems. Fuzzy logic, in a narrow sense, is a logical system, which is an extension of multivalued logic. The fuzzy logic works on a theory which relates to classes of objects with blurred boundaries in which membership is a matter of degree. This enables fuzzy systems applicable to broad range of parameters and expected output values in many aspects of science. The aim of this study was to create a fuzzy model, in the MATLAB environment, to aid physicians in interpreting the results of the microscopic urine analysis, considering the number of bacteria, RBC and WBC as well as turbidity of the sample.