원문정보
초록
영어
Breast imaging case not only image low level features but also has image semantic features. In order to implement multimode retrieval of breast imaging, using feature selection algorithm based on association rules to select features, digging out the associated relationship between image low level features and image semantic features, and then taking advantage of association classification algorithm to get image visual semantic features, which reduced the semantic gap between image low level features and visual semantic features, at last, making similarity measure combined with low level features, to make multimode retrieval come true. As the results show, this method improve the performance of breast imaging case retrieval and provide more meaningful decision support for doctors.
목차
1. Introduction
2. Establishment of Associative Classification Model
2.1. Related Concepts of Association Rules
2.2. Features Selection
2.3. Associative Classification Algorithm
2.4. Establishment of Breast Masses Shape Classification Model
2.5. Retrieval System
3. Retrieval Results
4. Conclusion
Acknowledgements
References