Diagnosis Technology Research Of Mammographic Masses in Content-based Image Retrieval
In order to assist doctor to diagnosis of mammo-graphicmasses, a method is proposed. 22 features are extractedfrom each queried region of interest (ROI).A k-nearest neighbor(KNN) algorithm is used to retrieve similar images fromdatabase, and further calculate the mutual information (MI)between the queried image and the images which are in theretrieval results, so as to improve the retrieval performance.Finally, the scheme takes the first nine images with the highestMI scores as the final retrieval results. With the purpose ofproviding available decision-making information of diagnosticaids, we compare and analyze three calculating methods ofdecision index. The Experiment results show that the method isbetter than method of using KNN only, and improve the accuracyof diagnosis effectively.
Song Li-xin Wang Qing-yan Wang Li
College of Electrical & Electronic Engineering Harbin University of Science and Technology Harbin , Dept.of gynecology Heilongjiang hospital Harbin , China
国际会议
成都
英文
1-4
2010-06-18（万方平台首次上网日期，不代表论文的发表时间）