A fuzzy decision support system for rheumatic fever in Nepal

To develop a decision support model for medical diagnosis is very complex due to the level of vagueness, complexity and uncertainty management, especially when the same symptoms indicate the multiple diseases as well as symptoms doesn’t reflect any accurate measurement value. In this paper we describe how fuzzy mathematical approach could be applied to develop a decision support model to diagnose of Rheumatic Fever in different stages in Nepal. The World Health Organization, World Heart Foundation, and Nepal Heart Foundation’s guidelines and information are considered for the background sources of the knowledge. A fuzzy triangular and trapezoidal membership function will be applied to determine the degree of symptom’s to ascertain the doctor’s belief of symptom’s severity. The integral input value will be assigned by a doctor based on his/her experiences and belief. These values will be used in the fuzzification process by applying the membership functions to obtain the degree (membership degree). Fuzzy inference mechanism will map entire rules with membership degree and fuzzy logical operators will be used to evaluate the strength of firing rules. Then aggregate the output of each rule and applied Mamdani’s Centre of Gravity (CoG) methods for defuzzification process.