Laser & Optoelectronics Progress, Volume. 61, Issue 18, 1811001(2024)

Deep-Learning-Based Self-Absorption Correction for Fan Beam X-Ray Fluorescence Computed Tomography

Mengying Sun1, Shanghai Jiang1、*, Xiangpeng Li1, Xin Huang1, Bin Tang1, Xinyu Hu1、**, Binbin Luo1, Shenghui Shi1, Mingfu Zhao1, and Mi Zhou2
Author Affiliations
  • 1Chongqing Key Laboratory of Optical Fiber Sensor and Photoelectric Detection, Chongqing University of Technology, Chongqing 400054, China
  • 2College of Science, Chongqing University of Technology, Chongqing 400054, China
  • show less
    Figures & Tables(15)
    Schematic diagram of the fan-beam XFCT system
    Projected sinograms. (a) Sinogram with absorption; (b) sinogram without absorption
    Some numerical simulation phantoms
    Schematic diagram of the neural network architecture
    Comparison of the projected sinograms. (a) (b) (c) Original sinograms; (d) (e) (f) corrected sinograms; (g) (h) (i) objective sinograms
    Comparison of reconstructed images of the projected sinograms. (a) (b) (c) Reconstructed original sinograms; (d) (e) (f) reconstructed corrected sinograms; (g) (h) (i) reconstructed objective sinograms
    Geant4 simulation imaging system
    Phantom parameters
    The incident X-ray energy spectrum
    Projection sinograms. (a) Uncorrected sinogram; (b) corrected sinogram
    Reconstructed images. (a) Uncorrected reconstructed image; (b) corrected reconstructed image
    The CNR values of the reconstructed images
    • Table 1. Dataset parameters

      View table

      Table 1. Dataset parameters

      ParameterValue
      Phantom componentGold nanoparticles
      Mass fraction /%0.3,0.6,0.9,1.2,1.5,1.8
      ShapeEllipses,rectangles
      Range of attenuation coefficients[0.5,5]
    • Table 2. Overall evaluation metrics of the test set

      View table

      Table 2. Overall evaluation metrics of the test set

      ParameterOriginal sinogramCorrected sinogramReconstructed uncorrected sinogramReconstructed corrected sinogram
      SSIM(mean±standard deviation)0.8083±0.12930.9587±0.01870.6795±0.23450.9127±0.0391

      NRMSE

      (mean±standard deviation)

      0.1445±0.07490.0410±0.01260.2365±0.13450.0602±0.0299
    • Table 3. Evaluation metrics between uncorrected image and corrected image

      View table

      Table 3. Evaluation metrics between uncorrected image and corrected image

      ParameterUncorrected imageCorrected image
      Average gradient20.342026.4239
      Standard deviation29.274037.6341
    Tools

    Get Citation

    Copy Citation Text

    Mengying Sun, Shanghai Jiang, Xiangpeng Li, Xin Huang, Bin Tang, Xinyu Hu, Binbin Luo, Shenghui Shi, Mingfu Zhao, Mi Zhou. Deep-Learning-Based Self-Absorption Correction for Fan Beam X-Ray Fluorescence Computed Tomography[J]. Laser & Optoelectronics Progress, 2024, 61(18): 1811001

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Imaging Systems

    Received: Dec. 29, 2023

    Accepted: Jan. 22, 2024

    Published Online: Sep. 14, 2024

    The Author Email: Shanghai Jiang (jiangshanghai@cqut.edu.cn), Xinyu Hu (hxy_dz@cqut.edu.cn)

    DOI:10.3788/LOP232787

    CSTR:32186.14.LOP232787

    Topics