The Study Of Osteoma Computer-aided Diagnosis Expert System

Osteoma is a hackneyed and hazardous orthopaedics disease, sorted by benign tumor, malignancy tumor, transfer carcinoma, cancroid affection, etc. and especially the teenager easily suffer from the kind of disease. This is the same as many kinds of tumours, the pathogeny of osteoma has not yet been known. Because of its indeterminacy and more frequently recrudescence of osteoma, the doctor may be confronted with a lot of difficulties in diagnosing osteoma and differentiating it's type. In the realm of biomedical engineering, the computer-aided medical expert system is a important topic, it is a good example that medicine combines with other subjects, such as computer science, engineering mathematics, cognize science, logic, psychology and so on. And it is applied to some medical domain, for example, internal medicine, gynecology, cytology and preventive medicine etc. and achieves favorable effect. To apply medical expert system technology in the fieldof the auxiliary diagnosis and its therapy of osteoma must largely improve the accuracy of osteoma diagnosis, it will bring the medical worker a great deal of conveniency and be beneficial for patients from suffering.This paper firstly reviewed the application of computer-aided medical expert system, analysed and probed into the involved key techniques in constructing an expert system. The third part of this paper , on the basis of discussing the characteristics of osteoma and it's clinic diagnose, combining with relation data model and the newest reseach result based on image content, proposed a frame knowledge representation based on image content. Theories and practices indicated that this method has the advantages of being easy to implement (using ralation database management system) and the ability of representing complex knowledge, Such as, semantic and image content knowledge etc. In succession, we discussed the core research topic梤easoning mechanism of medical expert system. By means of the analysis of diagnosis logic that the osteoma specialist carries out in practice and of the characteristics acquiring data from sufferer, we presented two kinds of reasoning model: analogy inference and statistics-fuzzy inference. The later uses statistics and fuzzy set theory, and partly possesses the machine learning ability, so that use statistics learning method to extract knowledge from raw database, following the increasing of the system's cases, and that gradually improves the system's diagnosis accuracy by adjusting the model's parameters. Our system would synthesize two kinds of methods' results. And then, we discussed the method of extractingknowledge from X-ray image and the method of evaluating system. The fourth part described the system structure, functionality, design idea and implement scheme, development tools of our osteoma computer-aided diagnosis expert system, especially, the logic structure of the raw-database and knowledge-base. The fifth part further discussed how to improve and expand our system, finally, we prospected the lookout of the computer's application in osteoma.The main new ideas in the paper are as follows: 1. Using ES technology for the diagnosis and treat of osteoma; 2. Applying the reaserch result for complex knowledge representation; 3. Proposing a frame knowledge representation's implementing method based on relation data model; 4. Presenting a statistics-fuzzy reasoning model with partly statistics self-learning ability.