Localized knowledge based system for diagnosis and treatment of pediatric pneumonia: the case of Tigrai Central Zone in Ethiopia

Abstract

The purpose of this work is to develop a localized knowledge-based system for diagnosis and treatment of pediatric pneumonia in Tigrai health institutions. The researcher also examined specific diagnostic case in Tigrai where customized knowledge-based system can play a role. The traditional method of medical diagnostic is characterized with the problem of precision and accuracy. The communicable diseases remain the most common health problems in the region. Most of the communicable disease incidence is attributed to low socioeconomic status, shortage of skilled manpower in the area, overcrowded number of patients and shortage of budget. In addition, lower practitioners like health extension workers, nurses and others are challenging using foreign language for diagnosis and treatment. Therefore, designing an efficient localized knowledge-based system for diagnosis and treatment of pediatric pneumonia would help institutions to minimize problems caused by lack of human experts and to improve the quality of medical services. In an effort to address such problem, this study attempted to design and develop a localized pediatric pneumonia knowledge-based system that can provide support for physicians and patients to facilitate the diagnosis and treatment of pediatric pneumonia patients. The required knowledge was acquired using both structured and unstructured interviews from domain experts who were selected using purposive sampling technique from St. Marry Hospital Axum, Axum University Referral Hospital and Axum Health Center. Relevant documents analysis method was also followed to capture explicit knowledge. Then, the acquired knowledge was modeled using decision tree that represents concepts and procedures involved in diagnosis and treatment of pediatric pneumonia. Production  rules were used to represent the domain knowledge and knowledge-based system was developed using Social-Wetenschappelijk Informatica (SWI) Prolog editor tool. In addition, local languages can be integrated with knowledge-based system to allow end-users to communicate with the system in a simpler and easier way. Moreover, in testing and evaluating of the prototype 24 (twenty-four) patients history are selected in order to test the accuracy of the prototype and also for ensuring whether the prototype satisfies the requirements of its end-users or not. Thus, the overall total performance of the prototype is 87.5%. The prototype achieved a good performance and met the objectives of the study.

Keywords: Knowledge, system, Localization, pediatric, pneumonia, swi prolog, Tigrai, health