Title :
A new multi-view identification scheme for breast masses in mammography using multi-agent algorithm
Author :
Sun, Li ; Xu, Weidong ; Li, Lihua ; Liu, Wei ; Zhang, Juan ; Shao, Guoliang
Author_Institution :
Inst. for Biomed. Eng. & Instrum., Hangzhou Dianzi Univ., Hangzhou, China
Abstract :
The classification of breast masses into benign and malignant categories plays an important role in the area of computer-aided diagnosis (CAD) of breast cancer. In this paper, in order to improve the accuracy and the robustness of the classification and reduce the false positive rates, we proposed one novel scheme that was based on information fusion in multi views. A series of contour and shape features of the masses were chosen, and new contour features were used. Then, a multi-classifier fusion based on multi-agent algorithm was introduced. And finally, a multi-view information fusion method was applied. Experimental results demonstrated that the proposed classification scheme achieved a higher accuracy than those schemes using the individual classifiers and the multi-classifier fusion technique in single-view.
Keywords :
cancer; computerised tomography; image classification; image fusion; mammography; medical image processing; multi-agent systems; breast cancer; breast masses classification; computer-aided diagnosis; false positive rate reduction; mammography; multiclassifier fusion technique; multiview identification scheme; multiview information fusion method; Benign tumors; Biomedical engineering; Breast cancer; Computer aided diagnosis; Instruments; Malignant tumors; Mammography; Multiagent systems; Shape; Sun; Mammogram; breast cases; information fusion; multi-agent; multi-view;
Conference_Titel :
Control and Automation (ICCA), 2010 8th IEEE International Conference on
Conference_Location :
Xiamen
Print_ISBN :
978-1-4244-5195-1
Electronic_ISBN :
1948-3449
DOI :
10.1109/ICCA.2010.5524436