A LINGUISTIC FUZZY EXPERT SYSTEM FOR CONTAGIOUS DISEASES DETECTION AND ISOLATION

M. S. Osigbemeh; F. O. Ogunwolu; A. A. Omoare; H. C. Inyiama

M. S. Osigbemeh Department of Electronics and Control Engineering, Nnamdi Azikiwe University, Awka
F. O. Ogunwolu Department of System Engineering, University of Lagos, Akoka, Lagos State
A. A. Omoare Department of Medical Microbiology and Parasitology, University of Ilorin, Kwara State
H. C. Inyiama Department of Electronics and Control Engineering, Nnamdi Azikiwe University, Awka

Keywords: Fuzzy Expert Systems, Contagious Diseases Control, SOSIC Diagnosis, Preventing Diseases, Quarantining

Abstract

This paper presents an electronic Expert system platform to detect and diagnose existing and new cases of contagious diseases as they occur with minimal contact with the index patient(s) and healthcare personnel with a confidence level that can be used to initiate or suggest appropriate follow-up action(s). The aim is to use ICT tools for patient-diagnosis, raise a red flag in real-time and thus contain contagious cases which may degenerate into an epidemic by providing a way to analyze vague and ambiguous input data from visible and reported symptoms in patients. A re-useable Expert system which makes use of fuzzy reasoning techniques and design methodology was used in this work. The Expert system is premised on rule-based fuzzy logic which captures the ambiguity, imprecision and nuances involved in disease reporting and detection using the Mamdani model. The software developed for the Fuzzy Expert system, called SOSIC, presents its diagnosis with fuzzy values between 0 to 1 corresponding to its level of confidence for the fuzzy inputs. The current approach to e-diagnosis and detection of contagious diseases using the SOSIC software is not completely contactless, thus ongoing investigations are geared towards improving SOSIC to be contactless. The developed system which runs on a computer system provides a safe procedure with minimum contact between patients and healthcare personnel to address early detection and diagnosis issues that may help forestall chain-infection and epidemics. The fuzzy based Expert system can be further extended to accommodate the detection of a wider array of symptoms as new cases arise; thus this paper fulfils an identified need in safe healthcare practice.