A cost-sensitive extension of AdaBoost with markov random field priors for automated segmentation of breast tumors in ultrasonic images

详细信息    查看全文

• 作者：Atsushi Takemura (1)
  Akinobu Shimizu (1)
  Kazuhiko Hamamoto (2)
  
• 关键词：Breast tumor ; Ultrasonic image ; Segmentation ; AdaBoost ; Cost ; sensitive extension
• 刊名：International Journal of Computer Assisted Radiology and Surgery
• 出版年：2010
• 出版时间：September 2010
• 年：2010
• 卷：5
• 期：5
• 页码：537-547
• 全文大小：967KB
• 参考文献：1. Shankar PM, Reid JM, Ortega H, Piccoli CW, Goldberg BB (1993) Use of non-Rayleigh statistics for the identification of tumors in ultrasonic B-scans of the breast. IEEE Trans Med Imaging 12(4): 687鈥?92 CrossRef
  2. Garra BS, Krasner BH, Horii SC, Ascher S, Mun SK, Zeman RK (1993) Improving the distinction between benign and malignant breast lesions: the value of sonographic texture analysis. Ultrason Imaging 15(4): 267鈥?85 CrossRef
  3. Haralick RM, Shanmugam K, Dinstein I (1973) Textural features for image classification. IEEE Trans Syst Man Cybern SMC- 3(6): 610鈥?21 CrossRef
  4. Weng L, Reid JM, Shankar PM, Soetanto K (1991) Ultrasound speckle analysis based on K-distribution. J Acoust Soc Am 89(6): 2992鈥?995 CrossRef
  5. Aleman-Flores M, Alvarez L, Caselles V (2007) Texture-oriented anisotropic filtering and geodesic active contours in breast tumor ultrasound segmentation. J Math Imaging 28: 81鈥?7 CrossRef
  6. Caselles V, Kimmel R, Sapiro G (1997) Geodesic active contour. Int J Comput Vision 22(1): 61鈥?9 CrossRef
  7. Shimizu A, Narihira T, Furukawa D, Kobatake H, Nawano S, Shinozaki K (2008) Ensemble segmentation using AdaBoost with application to liver lesion extraction from a CT volume. Proceedings of workshop in MICCAI2008
  8. Niessen W, Walsum T, Schaap M et聽al (2008) 3D segmentation in the clinic: a grand challenge II. Proceedings of workshop in MICCAI2008, http://grand-challenge2008.bigr.nl/
  9. Freund Y, Schapire RE (1997) A decision-theoretic generalization of on-line learning and an application to boosting. J Comput Syst Sci 55(1): 119鈥?39 CrossRef
  10. Besag J (1974) Spatial interaction and the statistical analysis of lattice systems. J Royal Statist Soc B 36(2): 192鈥?36
  11. Besag J (1986) On the statistical analysis of dirty pictures. J Royal Statist Soc B 48(3): 259鈥?02
  12. Perez P (1998) Markov random fields and images. CWI-Quarterly 11(4): 413鈥?37
  13. Masnadi-Shirazi H, Vasconcelos N (2007) Asymmetric boosting. Proceedings of the 24th international conference on machine learning, pp 609鈥?19
  14. Stan Z Li (2001) Markov random field modeling in image analysis. Springer, Berlin
  15. Nishii R, Eguchi S (2005) Supervised image classification by contextual AdaBoost based on posteriors in neighborhoods. IEEE Trans Geosci Remote Sens 43(11): 2547鈥?554 CrossRef
  16. Takemura A, Shimizu A, Hamamoto K (2010) Discrimination of breast tumors in ultrasonic images using an ensemble classifier based on the AdaBoost algorithm with feature selection. IEEE Trans Med Imaging. 29(3): 598鈥?09 CrossRef
  17. Perona P, Malik L (1990) Scale-spacing and edge detection using anisotropic diffusion. IEEE Trans Pattern Anal Machine Intell 12: 629鈥?39 CrossRef
  18. Press WH, Teukolsky SA, Vettering WT, Flannery BP (2003) Numerical recipes in C. Cambridge University Press, Cambridge, pp 683鈥?88
  19. Ibanez L, Schroeder W, Ng L, Cates J. The ITK software guide, http://www.itk.org/, pp 550鈥?55
  
• 作者单位：Atsushi Takemura (1)
  Akinobu Shimizu (1)
  Kazuhiko Hamamoto (2)
  
  1. Tokyo University of Agriculture and Technology, 2-24-16, Naka-cho, Koganei-shi, Tokyo, 184-8588, Japan
  2. Tokai University, Kanagawa, 259-1292, Japan
  

文摘

Purpose A cost-sensitive extension of AdaBoost based on Markov random field (MRF) priors was developed to train an ensemble segmentation process which can avoid irregular shape, isolated points and holes, leading to lower error rate. The method was applied to breast tumor segmentation in ultrasonic images. Methods A cost function was introduced into the AdaBoost algorithm that penalizes dissimilar adjacent labels in MRF regularization. The extended AdaBoost algorithm generates a series of weak segmentation processes by sequentially selecting a process whose error rate weighted by the cost is minimum. The method was tested by generation of an ensemble segmentation process for breast tumors in ultrasonic images. This was followed by a active contour to refine the extracted tumor boundary. Results The segmentation performance was evaluated by tenfold cross validation test, where 300 carcinomas, 50 fibroadenomas, and 50 cysts were used. The experimental results revealed that the error rate of the proposed ensemble segmentation was two-thirds the error rate of the segmentation trained by AdaBoost without MRF. By combining the ensemble segmentation with a geodesic active contour, the average Jaccard index between the extracted tumors and the manually segmented true regions was 93.41%, significantly higher than the conventional segmentation process. Conclusion A cost-sensitive extension of AdaBoost based on MRF priors provides an efficient and accurate means for the segmentation of tumors in breast ultrasound images.