Research And Application Of Medical Automatic Diagnosis System Based On Ontology

With the increasing of population,population aging is getting worse,and people's living standard is improving,people pay more attention to their health condition.At present,doctors are still using a combination of interrogation and medical tests to diagnose the disease,but the development of medical are not unified in different areas and different hospitals,doctor's qualifications are different at the same time,many patients may spend a lot of money for medical examination and diagnose.here are a lot of diagnostic errors.In this way,the key point of this article is how to realize automatically diagnose for patients.In the past 20 years,Ontology has been widely applied to knowledge engineering,artificial intelligence,information recommendation,natural language processing,bioinformatics,agriculture and other fields.There are more and more Ontology-based applications,however,ontology is hand-built in many cases.Their shortcomings are heavy workload,low efficiency and so on.This paper use ontology for knowledge representation,use the symptoms and disease knowledge to establishes the ontology knowledge base of disease and symptoms.It is convenient for automatic diagnosis.The main contents of the paper are as follows:1.For building ontology of diseases and symptoms,this paper improves attribute partial order structure graph algorithm,and builds disease ontology repository based on this algorithm.In order to characterize some symptoms of the disease,this paper proposes the concept of symptom information for disease data set.In the process of building attribute partial order structure graph,we select the set of common symptoms with the minimum amount of information.Experimental results show that use the common symptoms sets with the minimum amount of information for forming a partial order structure can reduce the redundancy of multiple symptoms.2.For disease data sets,we build the structure of ontology between diseases and symptoms,propose a model for the diagnosis of patients.In the diagnosis of patients,doctors can use the similarity between symptoms based on the ontolog structure and get the best match between the patient's symptoms and diseases.In the construction of the ontological structure of diseases and symptoms,body layers use training data set for training weights.Finally,the weighted average,calculated the patients' illness and similarity between specific diseases,and use the similarity to measure probability of patients suffering from the disease.Experimental results show that the accuracy of this algorithm for acute cystitis diagnosis and acute glomerulonephritis diagnosis is above 80%.3.This paper has designed and implemented an automatic medical diagnosis system.The system uses the MVC design pattern,and it is simple.The processes of medical diagnosis automatically are as follows: the doctor login s in the system and fill the patient information forms for the first step,then,the doctor should choose the way to diagnosis and identify patients' symptoms.Consequently,the system will automatically give the final results of diagnosis according to the patient's symptoms.The final results of diagnosis include a variety of possibility of diseases which the patient may have and the possibility of every disease.According to the tests,this system can easily carry out medical diagnosis for patients.