Data Mining Applications In The Medical Field

The purpose of medical data mining is to find out the potential and effective knowledge, information, model, relation, change from large amounts of medical data, to help doctors make diagnosis more rapid and accurate. This thesis mainly studies the association rule extraction and rough set and its application in the Medical images mining. Combine the features of medical images and the technical characteristics of the data mining, Achieved good results. The main work and innovations of this thesis are as follows:First of all, summarized the theoretical basis, system structure, major technique of data mining and did many researches on data preprocessing, feature selection of medical data mining. Summarized a technical route and process model of medical data mining by combining these two.On this basis, this thesis researched for medical data mining mainly from two aspects, first, did research on the association rules mining technology, Especially the Apriori algorithm and the gray level co-occurrence matrix, put forward a method for mining association rules base on the gray level co-occurrence matrix. And use this method to extraction association rule from the liver CT images of affiliated Hospital of Luzhou Medical College, also did the compare test of Apriori algorithm and its improved algorithm, to check their efficiency and classification ability. Second, Did research on the key technology of rough set mining, compare all kinds of algorithm of attribute reduction, study its application in the medical field, then use the rough set theory to analysis Brain tumor MRI images, get guide rules which of benefit for classification from the MRI features mining.