The Application Study Of Data Mining Technology In The Area Of Medical Imaging Data

In recent years,with the popularity of Radiology information systems(RIS)in the hospital,the system has accumulated a large number of examation records.How to utlize these data effectively,has become an urgent question concerned by many hospital and health authority leaders.This paper focused on the RIS data of two 3A large hospitals in the same region,here summarized the main research work as follows:1)The forecast of equipment examination amount of regional medical imaging based on gray markov model:Optimized for the inadequacies of the traditional gray prediction model,and then used grey markov to predict the amount of equipmentexamination.And the model is applied toforecasttheequipment examination amount of two 3A hospitals in the same region,then the results areanalyzed and compared.2)The automatic disease classification research onregional medical imaging examinationbased on support vector machine:To solve the problem ofthe automatic diseases classification of medical imaging results.Firstly,a small part of the data in the RIS database was used K-means to cluster ten disease categories;secondly,an optimization of SVM algorithm based on the particularity of medical texts was proposed,it greatly improving the classification accuracy;Finally,the optimized SVM automatic classification is applied to the two 3A hospitals in the same region,then medical characteristics oftwo hospitals according to disease classification results was analyzed.3)The repeat medical imaging examination analysis and associated miningfor two 3A hospital in the same region:(1)Analyzed the situation of patients transferbetween two hospitals;(2)Analyzed the proportion ofmedical image repetitive examations;(3)Studied the associationbetween patients transfer and the two hospital medical characteristics;(4)Studied the correlation between patient transfer and hospital imaging equipment;(5)Verified the situation of repetitive medical imaging examination by Apriori algorithm,analyzed the correlationbetween disease and the examination amount,and it effectively verified the former forecast results.The results of this study showed that patient choose hospital based on medicalcharacteristicsand the advanced imaging equipment,and different hospitals for imaging examination have different degrees of recognition,and conventional ultrasound examination have high repeatability,etc.This will help to assist the hospital leaders and health authorities to make a decision.