Abstract:
Many rural communities in Ethiopia have extremely limited access to medical advice. People travel long distances to clinics or medical facilities, and there is a shortage of medical experts in most of these facilities. This results in slow service, and patients end up waiting long hours without receiving any attention. Hence, a medical knowledge-based system can play a significant role in such cases where medical experts are not readily available.
This work presents the design and development of a knowledge-based system that aims to provide the Patients/physicians with medical advice and basic knowledge on the first three stages of kidney disease. This knowledge-based system typically takes a set of symptoms, treatment options and management of kidney as input and produce diagnostic advice as output.
In this study, we adopt the method of fact finding called knowledge acquisition which is a knowledge-based approach to extract facts using both structured and unstructured interviews from five domain experts which are selected using purposive sampling technique from Gondar university Hospital and secondary data is acquired from different sources, such as journal articles, health care guidelines, manuals, books and browsing different sites. The acquired knowledge from domain experts and document analysis is modeled using a decision tree. We included various sets of rules into our system by using production rule (If-Then-Action) representation techniques for detecting the first three stages of kidney disease. The prototype developed uses backward chaining to infer the rules and provide appropriate recommendations.
The data extracted from experts and documents are stored in the knowledge-base of the KBS. The graphical user interface, the KBS shell and the database of the model are developed by SWI-PROLOG. In this study, evaluation was carried out to determine the word accuracy using visual interactions through closed-ended and open-ended questioners and for system validation 15 test cases are selected to see the system was achieved the objectives of this study or not. Thus, based on the evaluators the overall total performance of the system is 84.5%. This work includes Self-Learning KBS technique which is tries to acquire the experience and knowledge automatically. As a final point, appropriate recommendations as future research directions in the area of knowledge based system are presented.
