Online Thai medical diagnostic system using case-based reasoning

Abstract

The paper proposes a case-based medical diagnostic system supporting queries in Thai language. The system is intended to reduce both expense and time required by doctor's visits by giving a preliminary diagnosis. It is a web based application established with PHP language and a MySQL database. MySQL is used in our system for storing previous diagnoses of cases entered by doctors. The users can query by inputting their symptoms in Thai. The inputs are used to select the best matches among prior cases in the database. We evaluated our proposed system with 33 queries for each of 50 diseases. The result showed a 67.87% accuracy of diagnoses, with a 99% confidence interval from 53.98 to 81.74%.