Research On Data Mining And Visualization Technique In Medical Information Management System

With the rapid development of the computer technology, the information technology permeates every industry area. The reform of medical system and sound of medical care system attach great importance to the hospital information construction based on large population, especially the policy and economic support in our country. The development of the hospital is increasingly enhanced, and the number of the hospital is increasing. The health care industry has considerable material conditions and talent reserves. At the same time, along with the construction and use of the hospital HIS system, the health care industry have accumulated large amounts of medical data, and the medical profession has research professionals, laid a certain foundation for data mining application. How to dig out the useful knowledge from a lot of business data in the hospital, how to apply it for the hospital management and clinical auxiliary diagnosis is a common problem facing many hospital administrators.The technology of data mining is the use of computer graphics and image technology, the theory is a technology of interactive processing, converting the data into graphics or images displayed on the screen. The data mining technology will be integrated into the visualization of medical information system data mining. Users for data mining has a more direct and intuitive clear understanding. In turn, it is provided effective method to active participation in the data mining process. At last, medical staffs provide more, faster, better service for patients by clinical consultation, diagnosis, clinical decision.Paper main research work is as follows:A method for evaluating medical treatment scheme using grey relational analysis is discussed and an optimum strategy for formulating treatment scheme is proposed.Hospital data strong influence of data mining is put forward, mainly uses the statistical diagnosis method- Cook distance method, find out the strong influence points, provide useful information for the diagnosis and treatment.Rough set theory to predict the medical disease evaluation method is put forward. Based on rough set decision and indicators system of ideas, prediction models, curative effect and evaluation model and comprehensive effect model are established to predict and evaluate the rationality of the medical methods.Contribution score dimension reduction method is put forward. Multidimensional data process to dimension reduction, through the principal component analysis and factor analysis, finding out the main factors affecting the analysis and the graphic, provide the important of effective interaction information.Graphic data mining to hospital information system is put forward. Combined with the hospital HIS system, mining information about drugs alarm, validity alarm and Automatic purchase plan, rational drug use and Policy-making management,etc. at the same time, achieve the hospital nursing management of patients information effective graphic.