Application Of Cluster Analysis Algorithm In Thalassemia Disease

With the continuous improvement of the degree of informatization in China,all kinds of information technology are developing rapidly and means have been widely used in the medical industry,the application of information systems can effectively record patient information,test results,diagnostic data and other information,if you can extract from this large amount of information to produce valuable laws or relationships,can provide support for diagnostic decision-making or scientific research assistance.As a kind of genetic disease,thalassemia is more common in two regions of our country,as the representative of regional diseases,whether in eugenics and eugenics testing,or in the detection of individual disease prevention,have a place.At present,the diagnosis of thalassemia disease is mainly genetic testing diagnosis,and accompanied by routine thalassemia screening testing,but genetic testing because of its high cost,testing time is longer,is not the clinical preferred test items,so the results of conventional thalassemia screening test on the need for further genetic testing plays a key role.In the routine thalassemia screening test,due to the large number of test indicators involved,comprehensive consideration is required to arrive at the final results.Traditional clinical medical diagnosis and treatment behavior is subject to personal experience and knowledge level,if it can automatically realize the comprehensive analysis and judgment of multi-indicators by means of information technology,and get the diagnosis results and classification,it will play a guiding role in further clinical diagnosis.This article is based on the strong clinical need for automated diagnosis of Thalassemia,in order to solve the problem of the lack of thalassemia automatic diagnosis function in the existing Lis System(laboratory information management system)of a third-class Guangxi Zhuang Autonomous Region Hospital,combining the characteristics of K-MEANS Algorithm,which is fast,sensitive to cluster center,and can automatically solve the classification problem,an automatic diagnosis solution for the disease is proposed,to reduce the rate of misdiagnosis and missed diagnosis caused by clinical dependence on personal experience and knowledge.The following studies have been carried out with a view to achieving automatic diagnosis of the disease:(1)Based on the study of clustering algorithms and the practical application of Thalassemia,aiming at the problem that the clustering algorithm combined Canopy algorithm with K-means Algorithm depends on the initial clustering center and is easy to get into the local optimal solution,the average difference degree method is improved,and the t-value selection method of Canopy algorithm is improved,further improve the operation speed of the Algorithm.Experiments with real thalassemia data from the hospital show that the proposed fusion clustering algorithm can improve the accuracy of classification results,and in the calculation of large data sets faster.(2)Applying the fusion clustering analysis algorithm proposed in this paper to the LIS system of the hospital,the automatic diagnosis function module of thalassemia disease is designed and realized.Through testing,the automated diagnosis function module for thalassemia can effectively improve the correct rate of diagnosis of thalassemia and provide automatic reminder of abnormal results.The results show that the automatic diagnosis function module of thalassemia can realize the automatic diagnosis of thalassemia disease and effectively reduce the chance of misdiagnosis and missed diagnosis caused by human factors.