DEVELOPING KNOWLEDGE BASED SYSTEM FOR DIAGNOSIS AND TREATMENT OF GASTRITIS DISEASE

Abstract

Gastritis is an inflammation of the stomach lining, which is fairly common and could have different causes. The gastritis diagnosis and treatment skill required by the physicians including clinical experience, good judgment, sense of humor, stress-tolerance, knowledge of available resources, problem solving ability and sense of anticipation. It is therefore important for individuals to have adequate knowledge of different types of gastritis disease to reduce the risk of the disease. Experts systems are rapidly growing technology and an active area of research in disease diagnosis and treatment. Expert systems use human expert knowledge to solve complex problems in many fields such as Health, science, engineering, business, and weather forecasting. In an effort to address such problem, this study attempts to design and develop a user friendly GUI knowledge-based system that can provide advice for physicians and patients to facilitate the diagnosis and treatment of pneumonic patients. To this end, knowledge is acquired using both structured and unstructured interviews from domain experts which are selected using purposive sampling technique from Addis Ababa Bethel Referral Hospital. Relevant documents analysis method is also followed to capture explicit knowledge. Then, the acquired knowledge is modeled using decision tree that represents concepts and procedures involved in diagnosis and treatment of pneumonia and production rule is used to represent the domain knowledge. The knowledge-based system is developed using the integration of two programing languages with the interfacing library of JPL. Knowledge is also extracted from domain experts and document analysis from gastritis treatment. Hence, the prototype of knowledge based system which provides gastritis patient diagnosis and treatment was developed using SWI-Prolog 8.2.3 and Java NetBeans IDE 8.2 with JDK 8 for integrating with graphical user interface respectively. The overall performance achieves 92.8% accuracy with an average system performance.