Issue 183, article 2

DOI:https://doi.org/10.15407/kvt183.01.025
Filatova Anna E., PhD (Engineering), Associate Professor of Computing Technology and Programming Department of National Technical University “Kharkiv Polytechnic Institute”, Frunze st., 21, Kharkov, 61002,

e-mail: filatova@gmail.com

SELECTION OF PARAMETERS OF THE VISUALIZATION QUALITY IMPROVEMENT METHOD OF MAMMOGRAMS. Kibernetika i vyčislitel’naâ tehnika, 2016, issue 183, pp.25-36.

Introduction. The main method of breast cancer diagnosis is X-ray mammography. Visualization quality of studied organs of during radiological examinations depends on many factors that are related both to characteristics of recording equipment (an energy of ionizing radiation, exposure time, spatial resolution, etc.) and to characteristics of a visualization object (thickness and density of the tissue, dimensions of anatomical structures, etc.). So the task of visualization quality improvement of mammograms due to digital image processing is an important scientific and practical task. To solve this problem, the author developed the visualization quality improvement method of mammograms that is called IMRI-MAM. The main idea of the IMRI-MAM method is reviewed in the article.

The purpose of the article is selection of the type and parameters of a sigmoid function for performing of nonlinear contrast enhancement of the IMRI-MAM method and a quality evaluation of processed images.

Results. In this article properties of different types of sigmoidal functions are examined. The sigmoidal function of exponential form is proposed to use in the IMRI-MAM method. The properties of exponential sigmoid function are investigated. Optimization of parameters of nonlinear function of contrast enhancement in the IMRI-MAM method is performed. Dependence of the parameters of the sigmoid function of nonlinear contrast enhancement from the statistical characteristics of processed image is shown. The analytical expressions for calculating the parameters of the sigmoid function are found. Brightness, contrast and completeness of brightness gradations are selected as parameters for assessing of image quality. Local criteria for assessing of the image quality are reviewed. An integral criterion of image quality based on the properties of reviewed local criteria is proposed.

Conclusions. Experimental verification of the IMRI-MAM method of the visualization quality improvement of mammograms with using exponential sigmoid function was performed. 350 mammograms obtained with the digital X-ray mammography complex SYMA (manufactured by “Radmir”, Kharkov, Ukraine) were processed using the IMRI-MAM method. Subjective assessment of the experts showed a significant quality improvement of the processed images, which is also confirmed by objective assessments of the images quality. Further studies are aimed at generalization of the IMRI-MAM method to handle different kinds of radiographic images.

Keywords: a sigmoid function, nonlinear contrast enhancement, mammogram, the IMRI-MAM method, a criterion of image quality.

Download full text (ru)!

References

  1. Ternovoj S.K., Abduraimov A.B. Radiation mammalogy. Moscow: GJeOTAR-Media, 2007, 128 p. (in Russian).
  2. Fisher U., Baum F. Mammography: 100 clinical cases. Moscow: MEDpress inform, 2009, 368 p. (in Russian).
  3. Bankman I. Handbook of Medical Image Processing and Analysis. London: Academic Press, 2008, 978 p.
  4. Gonzalez R., Woods R. Digital image processing. Moscow: Tehnosfera, 2005, 1072 p. (in Russian).
  5. Voronin V.V., Marchuk V.I. Methods and algorithms for image reconstruction with incomplete prior information: monograph. Shakhty: VPO «JuRGUJeS», 2010, 89 p. (in Russian).
  6. Krasil’nikov N.N. Digital processing of 2D- and 3D-image: Textbook. St. Petersburg: BHV-Petersburg, 2011, 608 p. (in Russian).
  7. Ezhova K.V. Modeling and Imaging . Textbook. St. Petersburg: NIU ITMO, 2011, 93 p. (in Russian).
  8. Jayaraman S. Digital Image Processing. Tata McGraw-Hill Education, 2011, 723 p.
  9. Gruzman I.S., Kirichuk V.S., Kosyh V.P. et.al. Digital image processing in information systems. Novosibirsk: Publishing house NGTU, 2002, 352 p. (in Russian).
  10. Solonina A.I., Klionskij D.M., Merkucheva T.V. et.al. Digital signal processing and MATLAB. St. Petersburg: BHV-Petersburg, 2013, 512 p.(in Russian).
  11. Eddins S. Image deblurring — Wiener filter (02.10.2007). Available at: http://blogs.mathworks.com/steve/2007/11/02/image-deblurring-wiener-filter.
  12. Bojko D.A., Filatova A.E. The method of visualization of pathological structures on mammograms using layerwise overlay. Vіsnik NTU «HPІ». Series: Mathematical modeling in engineering and technology. Kharkiv: NTU «KhPІ», 2014, № 6 (1049), pp. 29–34. (in Russian).
  13. D. Boyko, A. Filatova, Povoroznjuk A. International Conf. “Modern Problems of Radio Engineering, Telecommunications, and Computer Science”. The Imaging Method of Pathologic Structures on Mammograms Using Layerwise Overlay. Lviv Polytechnic National University, 2014, pp. 745–747.
  14. Mayne M. Ultimate Guide To Blending Modes in Photoshop. Available at: http://photodoto.com/how-to-master-blending-modes-in-photoshop/
  15. Ajsmann K., Palmer U. Retouching and image processing in Photoshop (3rd ed.). Moscow: Publishing House «Vil’jams», 2007, 560 p. (in Russian).
  16. Filatova A.E., Bojko D.A. The method of quality enhance of visualization of radiographic images. Vіsnik NTU «HPІ». Series: Information and simulation. Kharkiv: NTU «KhPІ», 2015, № 32 (1141), pp. 19–26. (in Russian).

Received 03.11.2015