Kidney failure diagnosis based on case-based reasoning (CBR) method and statistical analysis

Abstract

The kidney is one of the most important organs for human beings. It mainly functions to remove the waste products of the human body metabolism. 850.000 mortalities are caused by chronic kidney failure. According to the World Health Organization (WHO), chronic kidney failure was ranked as one of the top 12 causes of death in the world. For that reason, we need to develop CBR that can help in diagnosing kidney failure. CBR is a computer reasoning system which uses pre-existing cases and knowledge to solve new problems. CBR provides solutions to new cases by looking at the previous cases which are the most similar to new case. The patients' medical records on kidney failure are used as data. Calculation of similarity between the old and new cases was measured by using a simple matching coefficient. This study used a waterfall methodology begins with information system engineering, need analysis, design coding, and testing. For coding, authors used PHP programming language and MySQL data base. Having obtained the result of CBR statistics further tested whether CBR can be wholly accepted when there is a new case. The result of the CBR experiment showed that the system can diagnose kidney failure based on the experiment done by the experts with a 80% success rate and that rate has been tested by using a statistical Spearman rank test with a significant level of 5%, resulting there are 15 symptoms that can explain the stadium levels of the kidney failure patients.