Advertisement

Texture feature coding method for classification of liver sonography

  • Ming-Huwi Horng
  • Yung-Nien Sun
  • Xi-Zhang Lin
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1064)

Abstract

Liver sonography is a widely used noninvasive diagnostic tool. Analyzing histology changes in sonograms provides a means of diagnosing and monitoring chronic liver diseases. Nonetheless, conventional ultrasonography is still qualitative. To improve reliability of liver diagnosis, quantitative image analysis is highly desirable for the assessment of various liver states. In this paper, a novel approach, called Texture Feature Coding Method (TFCM) is presented for texture classification of liver sonography, more specifically, classification of normal liver, hepatitis and cirrhosis. TFCM is a texture analysis technique based on gray-level gradient variations in a 3×3 texture unit. It transforms an image into a texture feature image in which each pixel is represented by a texture feature number (TFN) coded by TFCM. The obtained texture feature numbers are then used to generate a TFN histogram and a TFN co-occurrence matrix which will produce texture feature descriptors. By coupling with a supervised maximum likelihood (ML) classifier, these descriptors form a classification system to discriminate the three above-mentioned liver classes. The TFCM-supervised ML system is trained by 30 liver samples proven by biopsy and tested on a set of 90 samples. The results show that the designed TFN-supervised ML system performs better than do existing techniques, and the correct classification rate can reach as high as 83.3%.

Keywords

Texture Feature Texture Classification Correct Classification Rate Liver Sonography Texture Unit 
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.

References

  1. 1.
    R.M. Haralick, K. Shanugan and I. Dinstein, “Texture features for image classification,” IEEE Trains. Syst. Man. Cybernet. Vol 3, pp. 610–621, 1973.Google Scholar
  2. 2.
    A. Duerinckx, K. Rosenberg, D. Aufrichting, A Beuget, G. Kanel and S. Lottenberg. “I” In vivio acoustic attention in liver correlation with blood tests and histology,” Ultrasound in Med. and Biol., vol. 14, pp 405–413, 1988.Google Scholar
  3. 3.
    Y.N. Sun, H.T. Chiu and X.Z. Lin, “A computer system for the analysis of liver cirrhosis from ultrasonic images,” Chinese Journal of Medical and Biological Engineering, Vol. 11, No. 2, pp. 119–135, 1991.Google Scholar
  4. 4.
    CM. Wu, Y.C. Chen, and K.S. Hsieh, “Texture feature for classification of ultrasonic liver images,” IEEE Trans. Med. Imaging, Vol. 11, No. 2, pp. 141–152, 1992.Google Scholar
  5. 5.
    C.M. Wu, and Y.C. Chen, “Statistical feature matrix for texture analysis,” Comput. Vision Graphic, Image Processing, Vol. 54, No. 5, pp 407–419, 1992.Google Scholar
  6. 6.
    D.C. He and L. Wang. “T”Texture features based on texture spectrum,” Pattern Recognition, Vol. 24, No. 5, pp.391–399, 1991.Google Scholar
  7. 7.
    D.C. He, L. Wang and J. Guibert, “Texture discrimination based on an optimal utilization of texture features,” Pattern Recognition, Vol. 21, No. 2, pp141–146, 1988.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1996

Authors and Affiliations

  • Ming-Huwi Horng
    • 1
  • Yung-Nien Sun
    • 1
  • Xi-Zhang Lin
    • 2
  1. 1.Institute of Information EngineeringNational Cheng Kung UniversityTainanTaiwan, R.O.C.
  2. 2.Department of Internal MedicineNational Cheng Kung UniversityTainanTaiwan, R.O.C.

Personalised recommendations