Statistical Characteristics of Portal Images and Their Influence in Noise Reduction

  • Antonio González-López
  • María-Consuelo Bastida-Jumilla
  • Jorge Larrey-Ruiz
  • Juan Morales-Sánchez
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7930)


Portal imaging is used in radiotherapy to asses the correct positioning of the patient before applying the treatment. Given the high energy particles used in portal image formation, portal image is intrinsically bound by low contrast and poor spatial resolution. The relevance of portal imaging in radiotherapy treatments and its common use justify efforts to improve its inherent low quality.

The knowledge of the statistical properties of both image and noise is essential in order to develop suitable processing algorithms to clean the image. The aim of this paper is to show how the statistical characteristics of the portal images and noise images generated in one of the portal imaging systems most widely deployed, can be exploited to improve the quality of noisy portal images through efficient denoising methods.

An ensemble of portal images is used to investigate their statistical characteristics. In the case of noise, a process of averaging and subtraction of the mean is used to extract noise images.

The distribution found for the noise is clearly Gaussian, in both the spatial and the wavelet domain. The curves for the noise show a parabolic shape in the semi-log graphs across the different scales, which translates into Gaussian character in the transformed domain. On the other hand, the probability density functions (pdf’s) for portal images show large tails.

Wavelet thresholding takes advantage of the different statistical features found for noise and signal. In the present work wavelet thresholding is compared to Wiener filtering, and the assesment of the denoised image is carried out by means of the peak signal to noise ratio PSNR and the structural similarity index SSMI.

Thresholding the wavelet coefficients of the noisy image gives better denoising results for both figures of merit (PSNR and SSIM) than the Wiener filter in all the analysed cases. Furthermore, the differences between the methods increase as the noise increases. abstract environment.


portal image image statistics wavelet processing denoising 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Herman, M., Balter, J., Jaffray, D., McGee, K., Munro, P., Shalev, S., Van Herk, M., Wong, J.: Clinical use of electronic portal imaging: Report of AAPM Radiation Therapy Committee Task Group 58. Med. Phy. 28(5), 712–737 (2001)CrossRefGoogle Scholar
  2. 2.
    Antonuk, L.: Electronic portal imaging devices: a review and historical perspective of contemporary technologies and research. Phys. Med. Biol. 47(6), R31 (2002)Google Scholar
  3. 3.
    Donoho, D., Johnstone, J.: Ideal spatial adaptation by wavelet shrinkage. Biometrika 81(3), 425–455 (1994)MathSciNetzbMATHCrossRefGoogle Scholar
  4. 4.
    Simoncelli, E., Adelson, E.: Noise Removal Via Bayesian Wavelet Coring (1996)Google Scholar
  5. 5.
    Portilla, J., Strela, V., Wainwright, M., Simoncelli, E.: Image denoising using scale mixtures of Gaussians in the wavelet domain. IEEE Trans. Image Process 12(11), 1338–1351 (2003)MathSciNetCrossRefGoogle Scholar
  6. 6.
    Ferrari, R.J., Winsor, R.: Digital radiographic image denoising via wavelet-based hidden Markov model estimation. Journal of Digital Imaging 18(2), 154–167 (2005)CrossRefGoogle Scholar
  7. 7.
    Weaver, J., Yansun, X., Healy, D., Cromwell, L.: Filtering noise from images with wavelet transforms. Magnetic Resonance in Medicine 21(2), 288–295 (1991)CrossRefGoogle Scholar
  8. 8.
    Donoho, D.: De-noising by soft-thresholding. IEEE Transactions on Information Theory 41(3), 613–627 (1995)MathSciNetzbMATHCrossRefGoogle Scholar
  9. 9.
    Donoho, D., Johnstone, J.: Adapting to Unknown Smoothness via Wavelet Shrinkage. Journal of the American Statistical Association 90, 1200–1224 (1995)MathSciNetzbMATHCrossRefGoogle Scholar
  10. 10.
    Vidakovic, B.: Nonlinear wavelet shrinkage with bayes rules and bayes factors. J. of the American Statistical Association 93, 173–179 (1998)MathSciNetzbMATHCrossRefGoogle Scholar
  11. 11.
    Stein, C.: Estimation of the Mean of a Multivariate Normal Distribution. The Annals of Statistics 9(6), 1135–1151 (1981)MathSciNetzbMATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Antonio González-López
    • 1
  • María-Consuelo Bastida-Jumilla
    • 2
  • Jorge Larrey-Ruiz
    • 2
  • Juan Morales-Sánchez
    • 2
  1. 1.Hospital Universitario Virgen de la ArrixacaMurciaSpain
  2. 2.Departamento de Tecnologías de la Información y las ComunicacionesUniversidad Politéctnica de CartagenaCartagenaSpain

Personalised recommendations