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

Laser & Optoelectronics Progress, Volume. 59, Issue 16, 1617004(2022)

Non-Rigid Registration Algorithm of Lung Computed Tomography Image Based on Multi-Scale Parallel Fully Convolutional Neural Network

Lihao Lin, Jianbing Yi*, Feng Cao, and Wangsheng Fang
Author Affiliations
  • School of Information Engineering, Jiangxi University of Science and Technology, Ganzhou 341000, Jiangxi , China
    • show less
    AbstractGet PDF(in Chinese)
    Figures&Tables (18)
    Equations (6)
    References (23)
    Cited By (7)
    Tools
    Paper Information
    Topics
    Figures & Tables(18)
    Flow chart of D-FCN lung image registration algorithm

    Fig. 1. Flow chart of D-FCN lung image registration algorithm

    Download full size
    Multi-scale parallel down-sampling module

    Fig. 2. Multi-scale parallel down-sampling module

    Download full size
    Lung feature maps at the same resolution and different scales. (a) Original image; (b) pooling core is 1; (c) pooling core is 2; (d) pooling core is 4

    Fig. 3. Lung feature maps at the same resolution and different scales. (a) Original image; (b) pooling core is 1; (c) pooling core is 2; (d) pooling core is 4

    Download full size
    Schematic diagram of dilated convolution. (a) Dilated factor is 1; (b) dilated factor is 2; (c) dilated factor is 3; (d) receptive fields with dilated factor of 1; (e) receptive fields with dilated factor of 2; (f) receptive fields with dilated factor of 3

    Fig. 4. Schematic diagram of dilated convolution. (a) Dilated factor is 1; (b) dilated factor is 2; (c) dilated factor is 3; (d) receptive fields with dilated factor of 1; (e) receptive fields with dilated factor of 2; (f) receptive fields with dilated factor of 3

    Download full size
    Pyramid dilated convolution module

    Fig. 5. Pyramid dilated convolution module

    Download full size
    Different feature maps at 1/8 resolution after three down sampling. (a) Large deformation feature map; (b) small deformation feature map

    Fig. 6. Different feature maps at 1/8 resolution after three down sampling. (a) Large deformation feature map; (b) small deformation feature map

    Download full size
    Adaptive channel attention module

    Fig. 7. Adaptive channel attention module

    Download full size
    Structure diagram of dilated FCN

    Fig. 8. Structure diagram of dilated FCN

    Download full size
    Original lung CT image and images under noise attack (axial surface). (a) Original image; (b) image with Gaussian noise; (c) image with salt and pepper noise

    Fig. 9. Original lung CT image and images under noise attack (axial surface). (a) Original image; (b) image with Gaussian noise; (c) image with salt and pepper noise

    Download full size
    Registration results of case1 in DIR-lab dataset (coronal plane). (a) Fixed image; (b) moving image; (c) deformed image; (d) difference before registration; (e) difference after registration

    Fig. 10. Registration results of case1 in DIR-lab dataset (coronal plane). (a) Fixed image; (b) moving image; (c) deformed image; (d) difference before registration; (e) difference after registration

    Download full size
    Box plot of target registration errors obtained by proposed algorithm from maximum inhalation to maximum exhalation in 10 cases in DIR-lab dataset

    Fig. 11. Box plot of target registration errors obtained by proposed algorithm from maximum inhalation to maximum exhalation in 10 cases in DIR-lab dataset

    Download full size
    Registration results of case1 in Creatis dataset (coronal plane). (a) Fixed image; (b) moving image; (c) deformed image; (d) difference before registration; (e) difference after registration

    Fig. 12. Registration results of case1 in Creatis dataset (coronal plane). (a) Fixed image; (b) moving image; (c) deformed image; (d) difference before registration; (e) difference after registration

    Download full size
    Box plot of target registration errors obtained by proposed algorithm from maximum inhalation to maximum exhalation in 6 cases of Creatis dataset

    Fig. 13. Box plot of target registration errors obtained by proposed algorithm from maximum inhalation to maximum exhalation in 6 cases of Creatis dataset

    Download full size

    • Table 1. Target registration errors (standard deviations) of 300 expert points in DIR-lab dataset for each improvement step of proposed algorithm

      View table

      Table 1. Target registration errors (standard deviations) of 300 expert points in DIR-lab dataset for each improvement step of proposed algorithm

      Multi-scale parallel

      down sampling

      		Pyramid dilated convolutionAdaptive channel attention

      Spatial

      regularization

      Data

      augmentation

      		Target registration error（standard deviation）/mm
      3.45（2.19）
      √2.77（1.77）
      √√2.60（1.97）
      √√√1.96（1.29）
      √√√√1.90（1.34）
      √√√√√1.71（1.20）

    • Table 2. Target registration errors (standard deviations) of 300 expert points in DIR-lab dataset with different dilated factors in module

      View table

      Table 2. Target registration errors (standard deviations) of 300 expert points in DIR-lab dataset with different dilated factors in module

      CaseDilated factors with 1，2，3，4Dilated factors with 1，2，4，8Dilated factors with 1，4，8，16Dilated factors with 1，8，16，24
      11.35（0.62）1.13（0.63）1.33（0.64）1.92（0.86）
      21.46（0.55）1.04（0.50）1.50（0.58）1.88（0.87）
      31.63（0.79）1.54（0.77）2.48（1.65）3.44（1.91）
      41.81（0.97）1.66（0.96）2.77（1.79）3.24（1.78）
      51.90（1.28）1.76（1.24）2.54（1.94）2.91（2.13）
      61.96（1.10）1.90（1.19）2.94（2.46）3.83（2.11）
      71.95（1.29）1.78（1.06）3.43（2.92）3.96（2.57）
      84.68（3.59）2.94（3.15）7.52（7.16）9.04（7.08）
      91.86（0.81）1.74（0.88）2.38（1.21）2.96（2.56）
      102.09（1.96）1.70（1.70）2.58（2.74）3.31（3.01）
      Mean2.06（1.29）1.71（1.20）2.95（2.31）3.64（2.48）

    • Table 3. Target registration errors (standard deviations) of 300 expert points in DIR-lab dataset of each registration algorithm

      View table

      Table 3. Target registration errors (standard deviations) of 300 expert points in DIR-lab dataset of each registration algorithm

      CaseInitialEppenhofR-NetDLIRFCND-FCN
      13.89（2.78）1.65（0.89）1.50（1.05）1.27（1.16）1.15（0.61）1.13（0.63）
      24.34（3.90）2.26（1.16）1.74（1.24）1.20（1.12）1.13（0.64）1.04（0.50）
      36.94（4.05）3.15（1.63）2.36（1.04）1.48（1.26）1.60（0.94）1.54（0.77）
      49.83（4.86）4.24（2.69）3.13（1.60）2.09（1.93）2.10（1.38）1.66（0.96）
      57.48（5.51）3.52（2.23）2.92（1.70）1.95（2.10）2.26（1.79）1.76（1.24）
      610.89（6.97）3.19（1.50）2.95（1.83）5.16（7.09）2.93（2.69）1.90（1.19）
      711.03（7.43）4.25（2.08）3.52（2.00）3.05（3.01）3.45（3.17）1.78（1.06）
      814.99（9.01）9.03（5.08）5.52（3.40）6.48（5.37）8.60（7.52）2.94（3.15）
      97.92（3.98）3.85（1.86）3.22（1.57）2.10（1.66）2.53（1.82）1.74（0.88）
      107.30（6.35）5.07（2.31）3.07（2.15）2.09（2.24）2.56（2.07）1.70（1.70）
      Mean8.46（6.58）4.02（3.08）2.94（1.80）2.64（4.32）2.83（3.67）1.71（1.20）

    • Table 4. Target registration errors (standard deviations) of 100 expert points in Creatis dataset of each registration algorithm

      View table

      Table 4. Target registration errors (standard deviations) of 100 expert points in Creatis dataset of each registration algorithm

      CaseInitialRPMFCND-FCN
      16.34（2.95）1.84（1.56）1.40（0.57）1.34（0.47）
      214.04（7.20）3.88（2.91）3.81（3.06）1.74（1.10）
      37.67（5.05）2.69（2.66）1.85（1.38）1.57（0.87）
      47.33（4.89）1.89（1.85）1.68（1.48）1.64（0.98）
      57.09（5.10）2.54（2.24）1.73（1.22）1.26（0.84）
      66.68（3.68）2.01（1.49）1.56（1.09）1.45（0.81）
      Mean8.15（5.60）2.47（2.27）2.01（1.46）1.50（0.85）

    • Table 5. Running time of each algorithm to register a set of lung CT images

      View table

      Table 5. Running time of each algorithm to register a set of lung CT images

      AlgorithmTime /s
      RPM900.00
      R-Net≈1.00
      DLIR0.30±0.05
      FCN0.63
      D-FCN0.10±0.05
    Tools

    Get Citation

    Copy Citation Text

    Lihao Lin, Jianbing Yi, Feng Cao, Wangsheng Fang. Non-Rigid Registration Algorithm of Lung Computed Tomography Image Based on Multi-Scale Parallel Fully Convolutional Neural Network[J]. Laser & Optoelectronics Progress, 2022, 59(16): 1617004

    Download Citation

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

    Category: Medical Optics and Biotechnology

    Received: Aug. 6, 2021

    Accepted: Sep. 24, 2021

    Published Online: Jul. 22, 2022

    The Author Email: Jianbing Yi (yijianbing8@163.com)

    DOI:10.3788/LOP202259.1617004

    Topics

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