2Optimization of Determinant Diagnostic Symptoms for Febrile Diseases using Genetic Algorithm

Abstract

Edith Ugochi Omede and Stella Chiemeke

Many diseases especially febrile diseases present numerous mimicking and confusing symptoms that pose great challenges to their proper distinctive syndromic diagnosis. This ambiguity causes inaccurate diagnoses which result in the misappropriation of treatment. Many victims of this situation have been left in worse health conditions or even death.This paper considers two febrile diseases that are believed to be in the blood of every Nigerian, these are Malaria and Typhoid Fever. This challenge of their distinctive syndromic diagnostic symptoms were tackled by optimizing the numerous symptoms using a genetic algorithm based on their manifestation degree (the frequency of occurrence of a symptom in different cases). The genetic algorithm was simulated using matlabR2013a. An optimization degree of 64.06% was obtained. Though the conventional method is the best for the disease diagnosis, it is not always available, especially in rural areas where many depend on low-skilled medical practitioners for their health care. The use of these optimized determinant symptoms in the syndromic diagnosis of Malaria and Typhoid fever will
reduce the risk of misdiagnosis of these two diseases.