| The popularity of information technology has led to the accumulation of abundant medical treatment data in various medical institutions.How to use these data to serve better service for the doctor's diagnosis has become the main field of medical research.Considering the uncertainty,incompleteness and redundancy of medical data,a medical assistant diagnosis system is designed and implemented based on rough set theory and evidence theory.The work of this thesis mainly includes the following aspects.(1)Analysis of the characteristics of medical treatment data.Using rough sets to classify patients with massive patient data(cases).A dynamic attribute reduction algorithm with rough entropy is proposed for the problems of dynamic disease diagnosis knowledge base and the problems in the reduction process.The algorithm corrects the original reduction relation on the basis of the new data and recalculates the rough entropy of the information according to the rough information of the original information,thus it can save the computation and improve the efficiency of the reduction.(2)Rough set theory and evidence theory combined with rule extraction.Based on the rough set theory and the theory of evidence,the decision rules are used to construct the diagnosis model.The decision rules in the decision table are used as the identification framework of the evidence theory,calculate the credibility and support of the rules,and use the evidence theory to make the decision results of different rules Evidence synthesis,and ultimately get the diagnosis.(3)System implementation and test.Based on the above method,the auxiliary diagnosis system was designed and implemented,and the function of the auxiliary diagnosis system was tested with rheumatism as an example.Comparison of the diagnosis of the system and the results of the original doctor's diagnosis,you can see that the system has achieved the effect of helping doctors diagnose and verified the effectiveness of the proposed method. |
PDF Full Text Request