Cancerous Liver Tissue Differentiation Using LVQ

  • Kuo-Sheng Cheng
  • Richard Sun
  • Nan-Haw Chow
Conference paper
Part of the Perspectives in Neural Computing book series (PERSPECT.NEURAL)


Medical image processing technique provides an objective and quantitative approach for characterizing the pathological tissue images. In this paper, the fractal dimension was used to quantify the image textures for differentiating the normal and cancerous cells in the liver tissue image. Several image enhancement methods and one edge detection method were applied for accentuating the objects of interest before fractal dimension estimation. From the results, it is shown that the edge-based histogram equalization would be the best one among all these image enhancement methods. In addition, any two fractal dimensions were combined as the input features to the learning vector quantization network for tissue classification. From the results obtained from ten normal and ten cancer cases, the accuracy was demonstrated to be more than 90%. Above all, the user-friendly graphical user interface was also developed in this study.


Fractal Dimension Gray Level Tissue Image Edge Detection Method Input Feature Vector 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    Public Health in Taiwan Area, Department of Health, The Executive Yuan, ROC, March 1995.Google Scholar
  2. [2]
    Erler BS, Troung HM, Kim SS, Huh MH, Geller SA, Marchevsky AM. A study of hepatocellular carcinoma using morphometric and densitometric image analysis. Am J Clin Pathol 1993; 100:151–157Google Scholar
  3. [3]
    Dougherty G. Quantitative indices for ranking the severity of hepatocellular carcinoma. Comput Med Imag & Graph 1995; 19:329–338CrossRefGoogle Scholar
  4. [4]
    Thiran JP, Macq B. Morphological feature extraction for the classification of digital images of cancerous tissues. IEEE Trans Biomed Eng 1996; 43:1011–1020CrossRefGoogle Scholar
  5. [5]
    Chow NH, Hsu PI, Lin XZ, et al. Expression of vascular endothelial growth factor (VEGF) in normal liver and hepatocellular carcinomas-An immunohistochemical study. Human Path 1997; 28:698–703CrossRefGoogle Scholar
  6. [6]
    Otsu N. A threshold selection method from gray-level histogram. IEEE Trans Syst Man Cybern 1979; 9:115–120CrossRefGoogle Scholar
  7. [7]
    Leu L. Image contrast enhancement based on the intensities of edge pixels. CVGIP: Graph Models and Image Proc 1992; 54:497–506CrossRefGoogle Scholar
  8. [8]
    Negrate AL, Beghdadi A, Dupoisot H. An image enhancement technique and its evaluation through bimodality analysis. CVGIP: Graph Models and Image Proc 1992; 54: 13–22CrossRefGoogle Scholar
  9. [9]
    Sarkar N, Chaudhuri BB. An efficient differential box-counting approach to compute fractal dimensions of image. IEEE Trans Syst Man Cybern 1994; 24:115–120CrossRefGoogle Scholar

Copyright information

© Springer-Verlag London 2000

Authors and Affiliations

  • Kuo-Sheng Cheng
    • 1
  • Richard Sun
    • 1
  • Nan-Haw Chow
    • 2
  1. 1.Institute of Biomedical EngineeringNational Cheng Kung UniversityTainanTaiwan, Roc.
  2. 2.Department of PathologyNational Cheng Kung University Hospital TainanTaiwan, ROC.

Personalised recommendations