Research Of Intelligent Medical Diagnostic Aided System Based On Fuzzy Neural Network

It is well known, the process of disease diagnosis is that doctors according to patient's clinical symptoms and their own medical knowledge to decide the cause of disease.Here refers the clinical symptom is broad sense, including the symptom,the body drafts,the experiment and all kinds of inspection,all that are the doctors'foundation to analyse the condition of disease and have a judgment about the disease.This diagnosis method is quite easily influenced.In order to help doctors considerate the diagnostic factors comprehensively,remove all of human's factors,so that to get accurately and objective diagnosis result, this article proposed an intelligent medical aided diagnosis system based on the fuzzy nerve network.This system can help doctors to have diagnostic result comprehensively and scientifically,make doctors keep away from take sickness as disease or when have a headache then cure head,avoid to diagnose easily.On the basis of medical expert's opinions and the suggestions, this thesis summarizes the abdominal pain disease knowledge, and utilized the fuzzy mathematic theory to process the clinical symptom knowledge and the case data. Main job as follows:1.Built the clinical symptoms knowledge base and the therapeutic schedule knowledge base on the basis of clinical abdominal pain disease and the disease diagnose experience.Analyze every kind of abdominal disease and the relevant of the causes of disease,diagnose ways and therapeutic schedules.2. According to abdominal pain diagnostic ways, with Delphi7.0 and SQL Server 2000 built an intelligence diagnosed system improve the diagnosing rightness rate.3. Built dozen nerve network model according to abdominal pain diagnostic ways, improved the speed of the neural networks reasoning.The source of the abdominal pain case data is come from a hospital of Liao Ning Province, and the data are processed. The corresponding neural networks have been established, and use this networks to study the case of abdominal pain knowledge, the simulation result indicated this system's diagnosis right rate reaches above 85%, conform to the requirement of the abdominal pain disease diagnosis, this result shows the system has superiority and can be put into effect.