Study on Cascade Classification in Abnormal Shadow Detection for Mammograms
Classifier plays an important role in a system detecting abnormal shadows from mammograms. In this paper, we propose the novel classification system that cascades four weak classifiers and a classifier ensemble to improve both computational cost and classification accuracy. The first several weak classifiers eliminate a large number of false positives in a short time which are easy to distinguish from abnormal regions, and the final classifier ensemble focuses on the remaining candidate regions difficult to classify, which results in high accuracy. We also show the experimental results using 2,564 mammograms.
KeywordsFeature Selection Classifier Ensemble Weak Classifier Fuji Photo Film Sequential Forward Selection
Unable to display preview. Download preview PDF.
- 2.Furuya, S., Wei, J., Hagihara, Y., Shimizu, A., Kobatake, H., Nawano, S.: Improvement of performance to discriminate malignant tumors from normal tissue on mammograms by feature selection and evaluation of feature selection criteria. Syst. and Comp. in Japan 35(5), 72–84 (2004)CrossRefGoogle Scholar
- 3.Nemoto, M., Shimizu, A., Kobatake, H., Takeo, H., Nawano, S.: Classifier ensemble for mammography CAD system combining feature selection with ensemble learning. In: Proc. 19th CARS, pp. 1047–1051 (2005)Google Scholar
- 4.Viola, P., Jones, M.: Rapid object detection using a boosted cascade of a simple feature. In: IEEE Computer Society Conf. on Computer Vision and Pattern Recognition (CVPR), vol. 1, pp. 511–518 (2001)Google Scholar
- 5.Wei, J., Hagihara, Y., Kobatake, H.: Detection of cancerous tumors on chest X-ray images – Candidate detection filter and its application. In: Proc. ICIP, no.27AP4.2 (1999)Google Scholar
- 6.Kass, M., Witkin, A., Terzopoulos, D.: SNAKES: Active contour models. In: Proc. 1st ICNN 1987, pp. 259–268 (1987)Google Scholar
- 7.Pudli, P., Ferri, F.J., Novovicova, J., Kittler, J.: Floating search methods for feature selection with nonmonotomic criterion functions. In: Proc. IEEE Int. Conf. on Pattern Recognition, pp. 279–283 (1994)Google Scholar
- 8.Nemoto, M., Shimizu, A., Hagihara, Y., Kobatake, H., Takeo, H., Nawano, S.: Improvement of tumor detection performance in mammograms by feature selection from a large number of features and proposal of fast feature selection method. Syst. and Comp. in Japan (to appear)Google Scholar
- 9.Doppman, J.L., Girton, M., Vermess, M.: The risk of hepatic artery embolization in the presence of obstructive jaundice. Radiology 143, 37–43 (1982)Google Scholar
- 10.Hanley, J.A., McNeil, B.J.: A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 148, 839–843 (1983)Google Scholar