Medical Image Segmentation by Level Set Method Incorporating Region and Boundary Statistical Information

  • Pan Lin
  • Chongxun Zheng
  • Yong Yang
  • Jianwen Gu
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3287)


Level set methods are powerful numerical techniques for image segmentation and analysis. This method requires the definition of a speed function that governs curve evolution. However, the classical method only used image gradient, edge strength, and region intensity to define the speed function. In this paper, we present a new speed function for level set framework. The new method integrates the image region statistical information and image boundary statistical information instead of the conventional method that uses spatial image gradient information. The new speed function gives the level set method a global view of the boundary information within the image. The method here proposed is particularly well adapted to situations where edges are weak and overlap, and images are noisy. A number of experiments on ultrasound, CT, and X-ray modalities medical images were performed to evaluate the new method. The experimental results demonstrate the reliability and efficiency of this new scheme.


  1. 1.
    Osher, S.J., Sethian, J.A.: Fronts propagating with curvature dependent speed: Algorithms based on Hamilton-Jacobi formulations. J. Comput. Phys. 79, 12–49 (1988)zbMATHCrossRefMathSciNetGoogle Scholar
  2. 2.
    Malladi, R., Sethian, J., Vemuri, B.: Shape modeling with front propagation: A level set approach. IEEE T-PAMI 17(2), 158–175 (1995)Google Scholar
  3. 3.
    Chakraborty, A., Staib, L., Duncan, J.: Deformable boundary finding in medical image by integrating gradient and region information. IEEE T-MI 15(6), 859–870 (1996)CrossRefGoogle Scholar
  4. 4.
    Ronfard, R.: Region-based strategies for active contour models. International Journal of Computer Vision 13(2), 229–251 (1994)CrossRefGoogle Scholar
  5. 5.
    Poon, C.S., Braun, M.: Image segmentation by a deformable contour model incorporating region analysis. Phys. Med. Biol. 42(9), 1833–1841 (1997)CrossRefGoogle Scholar
  6. 6.
    Xu, C., Jr. Yezzi, A.: J. Prince. On the relationship between parametric and geometric active contours. Technical Report JHU/ECE 99-14 (December 1999)Google Scholar
  7. 7.
    Baillard, C., Barillot, C.: Robust 3D segmentation of anatomical structures with level sets. In: Delp, S.L., DiGoia, A.M., Jaramaz, B. (eds.) MICCAI 2000. LNCS, vol. 1935, pp. 237–245. Springer, Heidelberg (2000)Google Scholar
  8. 8.
    Voorhees, H., Poggio, T.: Detecting textons and texture boundaries in natural images. In: Proc.of the International Conference on Computer Vision, pp. 250–258 (1987)Google Scholar
  9. 9.
    Geman, S., Geman, D.: Stochastic relaxation, Gibbs distributions and the Bayesian restoration of images. IEEE Trans. on Pattern Analysis and Machine Intelligence 6, 721–741 (1984)zbMATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2004

Authors and Affiliations

  • Pan Lin
    • 1
  • Chongxun Zheng
    • 1
  • Yong Yang
    • 1
  • Jianwen Gu
    • 1
  1. 1.Institute of Biomedical Engineering, Key Laboratory of Biomedical Information Engineering of Education MinistryXi’an Jiaotong Universityxi’anChina

Personalised recommendations