Expert System for Poisoning 


 Project

Abstract

A clinical decision support system (CDSS) tries to come up with a diagnosis and suggests recommendations in a short length of time depending on some patient's data, in our case, ingested substances, signs and symptoms related to poisoning. Applying CDSS on poisoning domain can compensate the lack of competent human resources and resource-intensive centers for poisoning in the Philippines. The implementation of this system would extend the reach of the medical experts even to the under-served regions. The project will be implemented as a rule-based expert system that would be built primarily with a knowledge base and an inferencing engine. Our CDSS would have a knowledge acquisition tool and a CLIPS-file converter, special components that will allow addition of knowledge into the knowledge base and create new decision rules that will be fed into the inferencing engine for generating results. For interoperability, the CDSS could be accessed through the web, with the help of the XML converter and some web services. In this paper, we describe the making of the knowledge base and the inferencing engine of our system; how we utilized the inferencing mechanism of CLIPS, which is an expert system shell; the algorithms used as well as the problems we've encountered; the results observed after testing the program; and also the possible solutions and adjustments we could make for the improvement and efficiency of our CDSS.

Methodology

Technologies Used:

  • JDK 1.5
  • JClips
  • GWT 1.4
  • MySQL Connector/J 5.1
  • Eclipse 3.2
  • Apache Tomcat 5.5.26
  • MySQL 5.0