Spline-based regularisation for discrete FBP reconstruction

  • J-P V Guedon
  • Y J Bizais
1. Image Formation And Reconstruction
Part of the Lecture Notes in Computer Science book series (LNCS, volume 511)


In this paper, we show that tomographic images are degraded by the unsuitable discretisation of continuous schemes, and the non-trivial null space in the case of angular sampling. Usually these two types of degradations are not studied separately. However, discretisation can be performed properly, while the null space is irreducible. For this reason, we study the relationships between continuous and discrete versions of a direct reconstruction method (FBP). They are characterized by an interpolation / sampling kernel, called the Pixel Intensity Distribution Model (PIDM). By defining the latter as B-spline functions, the existence and the uniqueness of the solution is guaranteed. It follows that projections must be oversampled. We test the robustness of this exact solution (for an infinite number of projections) by decreasing the number of angles. PIDM results are much better then FBP ones, showing that FBP reconstructed images are degraded not only by the null space, but also by unsuitable discretisation.

We also analyze the influence of degradations induced by an imaging device (mechanical instability and blur) and by projection noise in SPECT. Discretisation-related degradations depend on projection sampling. For this reason, proper oversampling is achieved when the corresponding degradations are negligible in comparison to the ones induced by the imaging device.

Our algorithm is constrained by the amount of information input to the system and controlled by the number of projection angles and the PIDM order. Optimal values of these parameters could be found for a predefined task using the ROC curve methodology.


B-splines Filtered Back-projection Pixel Intensity Distribution Model 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Radon J. (1917). Uber die Bestimmung von Functionen durch ihre Integralwerte langs gewisser Mannigfaltigkeiten. Berichte Sachsische Academie der Wissenchaften, Leipzig, Math. — Phys. Kl, Vol-69:262–267.Google Scholar
  2. Beylkin G. (1987). Discrete Radon Transform. IEEE Trans. Acoust., Speech Signal Proces. ASSP-35(2):162–172.Google Scholar
  3. Brooks R.A., Weiss G.H. and Talbert A.J. (1978). A new approach to interpolation in Computed Tomography. J. Comp. Ass. Tomography Vol. 2:577–585.Google Scholar
  4. Gordon R. (1974). A tutorial on ART. IEEE Trans. Nucl. Sci., NS-21:78–93.Google Scholar
  5. Guédon J-P. and Bizais Y. (1990). Projection and Backprojection models and projection sampling in Tomography. SPIE Medical Imaging IV, Vol 1231:206–217.Google Scholar
  6. Guédon J-P (1990). Sampling problems in Tomography, Ph.D. dissertation (in French), ENSM, Nantes.Google Scholar
  7. Guédon J-P. and Bizais Y. (1991). A Spline based tomographic reconstruction method. SPIE Medical Imaging V, Vol 1443.Google Scholar
  8. Hanson K. (1989). Optimisation for object localization of the constrained algebraic reconstruction technique. SPIE., Medical Imaging III, Vol 1090.Google Scholar
  9. Hanson K. (1989). Optimization of the constrained algebraic reconstruction technique for a variety of visual tasks. Proc. of the XIth IPMI Intern. Conference, Progress in clinical and biological research Vol. 363, Orthendhal & Llacer Eds., Wiley-Liss.Google Scholar
  10. Joseph P.M. (1982). An improved algorithm for reprojection rays through pixel images. IEEE Trans. on Med. Imaging, TMI-1:192–196.Google Scholar
  11. Katz M.B. (1978). Questions of uniqueness and resolution in reconstruction from projections. Lect. Notes Biomath. 26, Ed S. Levin, Springer-Verlag, Berlin.Google Scholar
  12. Louis A.K. (1981). Orthogonal function series expansions and the null-space of the Radon transform. medical image processing group, Technical report MIPG52, Buffalo, Univ. of New-York.Google Scholar
  13. Ludwig D. (1966). The Radon transform on Euclidean space. Comm. on pure and applied mathematics, Vol XIX:49–81.Google Scholar
  14. Myers K.J. and Hanson K.M. (1990). Comparison of the algebraic reconstruction technique with the maximum entropy reconstruction technique for a variety of detection tasks. SPIE Medical Imaging IV, Vol 1231:176–187.Google Scholar
  15. Myers K.J., Rolland J.P., Barrett H.H. and Wagner R.F. (1990). Aperture optimization for emission imaging: effect of a spatially varying background. JOSA, Vol. 7:1279–1293.Google Scholar
  16. Schmidlin P. and Doll J. (1989). Implementation of iterative reconstruction in PET. Proc. of the XIth IPMI Conference, Berkeley, CA, Progress in clinical and biological research. Vol. 363, Orthendhal & Llacer Eds., Wiley-Liss.Google Scholar
  17. Todd-Pokropek A. (1982). The mathematics and physics of emission computed tomography. Radionuclide Imaging 3–31.Google Scholar
  18. Unser M., Aldroubi A. and Eden M. (1990). A sampling Theory for Polynomial Splines. Intern. Symp. on Infor. Theory and its Appl. ISITA90:279–282.Google Scholar
  19. Unser M., Aldroubi A. and Eden M. (1991). Recursive Regularization Filters:Design, Properties, and Applications. IEEE trans. on pattern anal. mach. intel. PAMI-13-3.Google Scholar
  20. Wagner R.F. (1983). Low contrast sensitivity of radiologic, CT, nuclear medicine, and ultrasound medical imaging systems. IEEE Trans. on Medical Imaging, TMI-2:105–121.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1991

Authors and Affiliations

  • J-P V Guedon
    • 1
    • 2
  • Y J Bizais
    • 1
  1. 1.Imagerie Medicale Multimodalite, Faculte de Medecine de NantesFrance
  2. 2.Center for Devices and Radiological Health, FDARockvilleUSA

Personalised recommendations