Journals Articles Conferences News About CLP
Submit a paper Email Alert
Search
Search
Articles
Journals
News
Advanced Search
Top Searches
2D MaterialsTransformation opticsQuantum PhotonicsSmall lasersSemiconductor UV PhotonicsSupersymmetric microring laser arrays

Acta Optica Sinica, Volume. 39, Issue 8, 0811004(2019)

Low Dose Computed Tomography Image Reconstruction Based on Sparse Tensor Constraint

Jin Liu1,2,3, Yanqin Kang1,2, Yunbo Gu2,3,4, and Yang Chen2,3,4、*
Author Affiliations
  • 1 College of Computer and Information, Anhui Polytechnic University, Wuhu, Anhui 241000, China
  • 2 Laboratory of Image Science and Technology, Southeast University, Nanjing, Jiangsu 210096, China
  • 3 Key Laboratory of Computer Network and Information Integration (Southeast University),Ministry of Education, Nanjing, Jiangsu 210096, China
  • 4 School of Cyber Science and Engineering, Southeast University, Nanjing, Jiangsu 210096, China
    • show less
    AbstractGet PDF(in Chinese)
    Figures&Tables (10)
    Equations (0)
    References (19)
    Cited By (14)
    Tools
    Paper Information
    Topics
    Figures & Tables(10)
    Diagram of image block recombination

    Fig. 1. Diagram of image block recombination

    Download full size
    Diagram of tensor decomposition

    Fig. 2. Diagram of tensor decomposition

    Download full size
    Simulated reconstruction images of AAPM data (slice #60). (a1)-(a2) Results reconstructed by FBP and FCR methods for routine dose protocol; (b1)-(f1) results reconstructed by FBP, TV, GRR, FCR, and STCR methods for LD1 dose protocol; (b2)-(f2) results reconstructed by FBP, TV, GRR, FCR, and STCR methods for LD2 dose protocol

    Fig. 3. Simulated reconstruction images of AAPM data (slice #60). (a1)-(a2) Results reconstructed by FBP and FCR methods for routine dose protocol; (b1)-(f1) results reconstructed by FBP, TV, GRR, FCR, and STCR methods for LD1 dose protocol; (b2)-(f2) results reconstructed by FBP, TV, GRR, FCR, and STCR methods for LD2 dose protocol

    Download full size
    Simulated reconstruction images of AAPM data (slice #370). (a1)-(a2) Results reconstructed by FBP and FCR methods for routine dose protocol; (b1)-(f1) results reconstructed by FBP, TV, GRR, FCR, and STCR methods for LD1 dose protocol; (b2)-(f2) results reconstructed by FBP, TV, GRR, FCR, and STCR methods for LD2 dose protocol

    Fig. 4. Simulated reconstruction images of AAPM data (slice #370). (a1)-(a2) Results reconstructed by FBP and FCR methods for routine dose protocol; (b1)-(f1) results reconstructed by FBP, TV, GRR, FCR, and STCR methods for LD1 dose protocol; (b2)-(f2) results reconstructed by FBP, TV, GRR, FCR, and STCR methods for LD2 dose protocol

    Download full size
    Quantitative results. (a) PSNR of AAPM data; (b) SSIM of AAPM data; (c) CNR of AAPM data; (d) CNR of UIH data

    Fig. 5. Quantitative results. (a) PSNR of AAPM data; (b) SSIM of AAPM data; (c) CNR of AAPM data; (d) CNR of UIH data

    Download full size
    Reconstruction images of UIH data (slice #156, #271 and #466). (a1)-(d1) Results reconstructed by the FBP, TV, FCR, and STCR methods for slice #156; (a2)-(d2) results reconstructed by the FBP, TV, FCR, and STCR methods for slice #156; (a2)-(d3) results reconstructed by the FBP, TV, FCR, and STCR methods for slice #466

    Fig. 6. Reconstruction images of UIH data (slice #156, #271 and #466). (a1)-(d1) Results reconstructed by the FBP, TV, FCR, and STCR methods for slice #156; (a2)-(d2) results reconstructed by the FBP, TV, FCR, and STCR methods for slice #156; (a2)-(d3) results reconstructed by the FBP, TV, FCR, and STCR methods for slice #466

    Download full size
    Evaluation of different parameters of STCR algorithm for AAPM data. (a) PSNR and SSIM versus regularization parameter λ; (b) PSNR and SSIM versus Lagrange parameter β; (c) partial enlargement of reconstructed image under different λ and β; (d) PSNR and SSIM versus image block group size n×m

    Fig. 7. Evaluation of different parameters of STCR algorithm for AAPM data. (a) PSNR and SSIM versus regularization parameter λ; (b) PSNR and SSIM versus Lagrange parameter β; (c) partial enlargement of reconstructed image under different λ and β; (d) PSNR and SSIM versus image block group size n×m

    Download full size
    PSNR and SSIM versus iterations. (a) PSNR; (b) SSIM

    Fig. 8. PSNR and SSIM versus iterations. (a) PSNR; (b) SSIM

    Download full size

    • Table 1. Parameters for data scanning in experiment

      View table

      Table 1. Parameters for data scanning in experiment

      DataAAPMUIH
      Distance from source to detector /cm108.56106.23
      Distance from source to isocenter /cm59.557.0
      Projection view11521200
      Tube voltage /kV100120
      Tube current /mA360(RDCT)20(LDCT)
      Detector size736×64(1.2856 mm×1.0947 mm)936×80(1.548 mm×1.405 mm)
      Reconstruction image size512×512(0.74 mm×0.74 mm)512×512(0.7828 mm×0.7828 mm)

    • Table 2. Computation time of different methods

      View table

      Table 2. Computation time of different methods

      MethodCalculation time /s
      AAPMUIH
      FBP1.02±0.201.63±0.20
      TV12.0±2.119.0±2.7
      GRR16.0±3.426.0±3.7
      FCR36.0±4.357.0±4.5
      STCR21.0±3.634.0±3.8
    Tools

    Get Citation

    Copy Citation Text

    Jin Liu, Yanqin Kang, Yunbo Gu, Yang Chen. Low Dose Computed Tomography Image Reconstruction Based on Sparse Tensor Constraint[J]. Acta Optica Sinica, 2019, 39(8): 0811004

    Download Citation

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

    Category: Imaging Systems

    Received: Mar. 15, 2019

    Accepted: May. 5, 2019

    Published Online: Aug. 7, 2019

    The Author Email:

    DOI:10.3788/AOS201939.0811004

    Topics

    laser devices and laser physics
    Lasers and Laser Optics
    Laser physics
    laser manufacturing
    Instrumentation, Measurement and Metrology