Physics in Medicine and Biology · 1994 · 53 citations · 9 references
ComputingComputer ScienceBiomedical ImagingComputational GeometryDigital Imaging3D Imaging3D ReconstructionComputed TomographyImage ReconstructionThree-dimensional ReconstructionComputed Tomography ReconstructionCone-beam Computed Tomography ReconstructionLine IntegralsProjection TheoryRadonReconstruction AlgorithmsThree-dimensional Image ReconstructionTomographic Reconstruction
We have used direct Fourier techniques to modify and implement the 3D reconstruction method from cone-beam projections proposed by Grangeat. In this way we manage to decrease the computational complexity from O(N4) to O(N3 log N). Just as Grangeat's original method is exact in the mathematical sense, so is our method, provided a complete set of projection data is acquired. Also in accordance with Grangeat, our algorithm consists of two distinct phases: phase 1, from cone-beam data to derivatives of Radon data; phase 2, from derivatives of Radon data to reconstructed 3D object. In phase 1 we use the direct Fourier method in reverse to obtain line integrals in the detector plane. In phase 2 the 2D linogram method is employed for reconstruction of vertical and horizontal planes in the Radon space.
9
L.A. Feldkamp, L. C. Davis, J. W. Kress · Journal of the Optical Society of America A · 1984 · 6.2K citations
An Inversion Formula for Cone-Beam Reconstruction
Heang K. Tuy · SIAM Journal on Applied Mathematics · 1983 · 933 citations
Engineering, Radiology, Optics +12