상세정보
학술지
A Sequential Feature Filtering Approach for Breast Magnetic Resonance Image Similarity Study
Cited 0 time in 
Download 1 time
Share 
Download 1 time
Share
- 저자
- Byung K. Jung, Wei Wang, Sung Y. Shin, 최형도, 백정기
- 발행일
- 201308
- 출처
- Information : An International Interdisciplinary Journal, v.16 no.8(B), pp.6259-6268
- ISSN
- 1343-4500
- 출판사
- International Information Institute
- 협약과제
- 13PR2600, 전자파 이용 조기진단 고정밀 MT 시스템 개발, 전순익
- 초록
- In this paper, we propose a new breast Magnetic Resonant Image (MRI) retrieval method based on tumor shape feature similarity. In this proposed approach we assume the images are classified into single objects through other known classif챠cat챠on methods such as K-means or Support Vector Machine (SVM) algorithms. From collected binary images of tumor objects, we develop a new algorithm that has less computation but equal accuracy as using shape feature - the curvature of the contour. We have experimented with classified binary tumor object images 첫om actual breast MRIs used in real medical diagnosis. Actual experimental results show that the proposed algorithm achieves equal retrieval results against traditional image retrieval using curvature of the contour with higher efficiency. © 2013 International Information Institute.
- KSP 제안 키워드
- Feature filtering, Feature similarity, Higher efficiency, Image retrieval, Image similarity, Magnetic resonance(MR), Magnetic resonance images, Magnetic resonant, Medical diagnosis, Single objects, Support VectorMachine(SVM)