Designing an Intelligent Feedback-based Medical Diagnosing andTreatment System

Data Mining has been helping Medical Science for decades to decrease the rate of mistaken diagnosis and treatment. Heart disease as the maincause of death around the world has been attracted a lot of attention by medical decision-support system designers. Almost all of the works inthis domain has focused on finding the most relevant factors of heart disease.In this paper we have proposed a diagnosis and treatment system with three main modules: Diagnosis module, Treatment module and Accuracy module. Diagnosis module builds a Bayesian Network for diagnosing, then Treatment Module uses KNN to find the best treatment based on the patient attributes. Accuracy module is our main contribution that uses sensitivity analysis to find the most costprofitable medical experiments for making diagnosis more accurate. We have tested our system on Data gathered from CCU ward of Namazi Hospital at Shiraz. The results confirm that our system is promising in all three parts

Keywords:

Medical Decision-Support System , Data Mining , Medical Diagnosisand Treatment , Heart Disease