A Novel Method for Predicting Pixel Value Distribution Non-uniformity Due to Heel Effect of X-ray Tube in Industrial Digital Radiography Using Artificial Neural Network

  • E. Nazemi
  • A. Movafeghi
  • B. RokrokEmail author
  • M. H. Choopan Dastjerdi


The heel effect in X-ray radiation imaging systems causes a non-uniform radiation distribution on the imaging plane. In this research, a novel method was developed for predicting the pixel value’s non-uniformity due to the heel effect of X-ray tube on the imaging plane using Monte Carlo N particle (MCNP) simulation code and artificial neural network (ANN). At first, an industrial X-ray tube and a computed radiography (CR) image plate were simulated using MCNP. Then, the simulation procedure was benchmarked with an experiment. In the next step, nine images were obtained from the simulation for nine different tube voltages in the range of 100–300 kV. Furthermore, some pixels with tangential and polar angles in the range of 0°–20° and of 0°–180° with respect to the centered pixel were chosen from these nine simulated images in order to train the ANN, respectively. The tube voltage, tangential and polar angles of each pixel were used as the three inputs of the ANN and gray value in each pixel was used as the output. After training, the proposed ANN model could predict the gray value of each pixel on the imaging plane with mean relative error of less than 0.23%. In the last step, the predicted gray value difference between the centered pixel and other pixels was calculated. The great advantage of proposed methodology is providing the possibility of predicting the pixel value’s non-uniformity due to the heel effect of the X-ray tube on the imaging plane for a wide range of the tube voltages and source to film distances independent of the tube current and exposure time. Although the proposed methodology in this paper was developed for a specific X-ray tube and a CR imaging plate, it can be used for every digital radiography system.


Heel effect X-ray tube Industrial digital radiography Monte Carlo simulation Artificial neural network 


  1. 1.
    Nascimento, M.Z., Frère, A.F., Germano, F.: An automatic correction method for the heel effect in digitized mammography images. J. Digit. Imaging 21, 177–187 (2008)CrossRefGoogle Scholar
  2. 2.
    Pawluczyk, O., Yaffe, M.J.: Field nonuniformity correction for quantitative analysis of digitized mammograms. Med. Phys. 28(4), 438–444 (2001)CrossRefGoogle Scholar
  3. 3.
    Behiels, G., Maes, F., Vandermeulen, D., Suetens, P.: Retrospective correction of the heel effect in hand radiographs. Med. Image Anal. 6, 183–190 (2000)CrossRefGoogle Scholar
  4. 4.
    Yu, Y., Wang, J.: Heel effect adaptive flat field correction of digital X-ray detectors. Med. Phys. 40(8), 1–10 (2013)CrossRefGoogle Scholar
  5. 5.
    Pelowitz, D.B.: MCNP-X TM User’s Manual, Version 2.5.0. LA-CP-05e0369. National Laboratory, Los Alamos (2005)Google Scholar
  6. 6.
    IEC 60336: Medical Electrical Equipment—X-ray Tube Assemblies for Medical Diagnosis—Characteristics of Focal Spots. IEC, Washington, DC (2005)Google Scholar
  7. 7.
    Boone, J.M., Seibert, J.A.: An accurate method for computer-generating tungsten anode X-ray spectra from 30 to 140 kV. Med. Phys. 24, 1661–1670 (1997)CrossRefGoogle Scholar
  8. 8.
    Rowlands, J.A.: The physics of computed radiography. Phys. Med. Biol. 47, 123–166 (2002)CrossRefGoogle Scholar
  9. 9.
    Baniukiewicz, P.: Automated defect recognition and identification in digital radiography. J. Nondestruct. Eval. 33, 327 (2014)CrossRefGoogle Scholar
  10. 10.
    Karami, A., Roshani, G.H., Salehizadeh, A., Nazemi, E.: The Fuzzy logic application in volume fractions prediction of the annular three-phase flows. J. Nondestruct. Eval. 36, 35 (2017)CrossRefGoogle Scholar
  11. 11.
    Tansel, I.N., Inanc, F.: Neural network based thickness estimation from multiple radiographic images. J. Nondestruct. Eval. 25, 53 (2006)CrossRefGoogle Scholar
  12. 12.
    Zapata, J., Vilar, R., Ruiz, R.: Automatic inspection system of welding radiographic images based on ANN under a regularisation process. J. Nondestruct. Eval. 31, 34 (2012)CrossRefGoogle Scholar
  13. 13.
    Felisberto, M.K., Lopes, H.S., Centeno, T.M., Arruda, L.V.R.: An object detection and recognition system for weld bead extraction from digital radiographs. Comput. Vis. Image Underst. 102, 238–249 (2006)CrossRefGoogle Scholar
  14. 14.
    El-Tokhy, M.S., Mahmoud, I.I.: Classification of welding flaws in gamma radiography images based on multi-scale wavelet packet feature extraction using support vector machine. J. Nondestruct. Eval. 34, 34 (2015)CrossRefGoogle Scholar
  15. 15.
    Zapata, J., Vilar, R., Ruiz, R.: An adaptive-network-based fuzzy inference system for classification of welding defects. NDT&E Int. 43, 191–199 (2010)CrossRefGoogle Scholar
  16. 16.
    Vilar, R., Zapata, J., Ruiz, R.: An automatic system of classification of weld defects in radiographic images. NDT&E Int. 42, 467–476 (2010)CrossRefGoogle Scholar
  17. 17.
    Boaretto, N., Centeno, T.M.: Automated detection of welding defects in pipelines from radiographic images DWDI. NDT&E Int. 86, 7–13 (2017)CrossRefGoogle Scholar
  18. 18.
    Sikora, R., Baniukiewicz, P., Chady, T., Łopato, P., Piekarczyk, B., Psuj, G., Grzywacz, B., Misztal, L.: Detection and classification of weld defects in industrial radiography with use of advanced AI methods. In: Far East forum on nondestructive evaluation/testing: new technology and application, 17–20 June 2013, Jinan, ChinaGoogle Scholar
  19. 19.
    Taylor, J.G.: Neural networks and their applications. Wiley, West Sussex (1996)Google Scholar
  20. 20.
    Gallant, A.R., White, H.: On learning the derivatives of an unknown mapping with multilayer feed forward networks. Neural Netw. 5, 129–138 (1992)CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Nuclear Science and Technology Research InstituteTehranIran

Personalised recommendations