Research Of Disease Diagnosis Based On The Hybrid Reasoning Technology

The doctor diagnoses the disease of patients by using clinical medical knowledge and theaccumulation of experience of many years, and obtains corresponding data through inquiringsymptoms, assay test, B-mode ultrasound, nuclear magnetic resonance and other medicalprocedures, then judges and infers what disease the patient has and which drugs should be usedfor treatment. In this paper, it simulates the reasoning process of doctors by intelligentreasoning methods, which overcomes the shortcomings of inefficiency of reasoning that causedby difficult to obtain incremental acquisition of complex diseases information based on ruleinference. At the same time, it uses the rule reasoning results as the index in the medical recordretrieval, which greatly improves the search speed.Firstly, it format and arrange the case data which accumulated over the years. It use theoptimized decision tree algorithm to select nodes, to carry contribution and to extract rulesaccording to the importance degree of disease feature attribute, and establish rule diagnosisknowledge base; Secondly, according to made fuzzy mathematics and nearest neighboralgorithm based on the characteristics of disease data retrieval hybrid algorithm; Finally, itquery the data case database according to the patient’s condition and rule reasoning results, andfind out the closest case as the important basis for diagnosis.For the above reasoning machine with integrated machine learning methods and theinternal implementation of primary and secondary to the main rule-based reasoning, case-basedreasoning, supplemented by an integrated strategy to carried out a mathematical modeling ofhybrid inference engine, computer program design and coding. And it achieved the rapid andaccurate diagnosis target through the simulation of more than1000cases.