Laser & Optoelectronics Progress, Volume. 61, Issue 8, 0837008(2024)

Projection Domain Denoising Method for Multi-Energy Computed Tomography via Dual-Stream Transformer

Shunxin Ouyang1, Zaifeng Shi1,3、*, Fanning Kong1, Lili Zhang1, and Qingjie Cao2
Author Affiliations
  • 1School of Microelectronics, Tianjin University, Tianjin 300072, China
  • 2College of Mathematical Science, Tianjin Normal University, Tianjin 300387, China
  • 3Tianjin Key Laboratory of Imaging and Sensing Microelectronics Technology, Tianjin 300072, China
  • show less
    Figures & Tables(11)
    Structure diagram of the proposed DTE-PDN-based low-dose multi-energy CT denoising method
    Structure diagram of the dual-stream Transformer
    Loss convergence curves on simulation dataset
    Comparison of denoising effects of abdomen slice from different methods
    Comparison of denoising effects of chest slice from different methods
    Ablation experiment results of different loss function combinations. (a) Selected projection path; (b) linear attenuation coefficient curves
    Curves of the evaluation index of the chest slice processed by DTE-PDN trained by the hybrid loss with different weight combinations
    • Table 1. Evaluation index of abdomen slice test results from different methods

      View table

      Table 1. Evaluation index of abdomen slice test results from different methods

      Energy /keVIndexRICCS-SI22RED-CNN23DRN9TransCT24DTE-PDN
      20‒80PSNR /dB33.941336.477337.006737.502237.8992
      SSIM0.96920.98520.98930.99500.9958
      FSIM0.92140.95150.97420.99630.9972
      80‒100PSNR /dB34.381537.792738.099338.968139.2680
      SSIM0.97030.98830.98910.99750.9964
      FSIM0.92570.95430.97600.99690.9976
      100‒120PSNR /dB35.016438.736239.095739.515339.6702
      SSIM0.97550.99010.99330.99750.9977
      FSIM0.92480.95560.97700.99710.9971
    • Table 2. Evaluation index of chest slice test results from different methods

      View table

      Table 2. Evaluation index of chest slice test results from different methods

      Energy /keVIndexRICCS-SI22RED-CNN23DRN9TransCT24DTE-PDN
      20‒80PSNR /dB34.886136.978337.346137.532337.7301
      SSIM0.97200.98730.98950.99270.9944
      FSIM0.93110.95500.97620.99770.9976
      80‒100PSNR /dB36.491338.287438.682039.675739.7969
      SSIM0.97670.98920.99120.99560.9961
      FSIM0.93680.95680.97740.99770.9981
      100‒120PSNR /dB37.375139.203139.686839.852939.8615
      SSIM0.97930.99060.99310.99810.9984
      FSIM0.93250.95660.97710.99830.9984
    • Table 3. Comparison of the mean and standard deviation among ROIs in test results from different methods

      View table

      Table 3. Comparison of the mean and standard deviation among ROIs in test results from different methods

      ROIEnergy /keVRICCS-SI22RED-CNN23DRN9TransCT24DTE-PDN
      ROI A20‒800.2020±0.00420.2021±0.00300.2022±0.00270.2024±0.00190.2028±0.0017
      80‒1000.1771±0.00340.1769±0.00230.1770±0.00200.1773±0.00150.1770±0.0015
      100‒1200.1672±0.00300.1665±0.00210.1680±0.00180.1678±0.00140.1677±0.0013
      ROI B20‒800.2116±0.00430.2116±0.00310.2115±0.00320.2116±0.00280.2117±0.0028
      80‒1000.1829±0.00360.1826±0.00270.1827±0.00260.1830±0.00230.1829±0.0022
      100‒1200.1728±0.00330.1728±0.00260.1731±0.00240.1732±0.00210.1730±0.0020
    • Table 4. Evaluation index of abdomen and chest slices from DTE-PDN with different combination of blocks

      View table

      Table 4. Evaluation index of abdomen and chest slices from DTE-PDN with different combination of blocks

      SliceSTULTU

      Residual

      structure

      20‒80 keV80‒100 keV100‒120 keV
      PSNR /dBSSIMFSIMPSNR /dBSSIMFSIMPSNR /dBSSIMFSIM
      Abdomen37.48530.98710.979638.85440.98470.980339.17810.98550.9814
      37.49650.98720.990838.95680.98790.985639.25170.98670.9823
      37.60300.99220.995339.19610.99310.993939.51800.99190.9948
      37.89920.99640.997239.26800.99640.997239.67020.99770.9971
      Chest37.31120.98610.978839.02170.98760.979039.11710.98730.9819
      37.40830.98730.985439.14450.99080.983539.27180.98950.9831
      37.48460.99070.995239.60010.99290.995339.67920.99230.9954
      37.73010.99440.997639.79690.99610.998139.86150.99840.9984
    Tools

    Get Citation

    Copy Citation Text

    Shunxin Ouyang, Zaifeng Shi, Fanning Kong, Lili Zhang, Qingjie Cao. Projection Domain Denoising Method for Multi-Energy Computed Tomography via Dual-Stream Transformer[J]. Laser & Optoelectronics Progress, 2024, 61(8): 0837008

    Download Citation

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

    Category: Digital Image Processing

    Received: Jun. 1, 2023

    Accepted: Aug. 8, 2023

    Published Online: Mar. 13, 2024

    The Author Email: Shi Zaifeng (shizaifeng@tju.edu.cn)

    DOI:10.3788/LOP231426

    Topics