Laser & Optoelectronics Progress, Volume. 59, Issue 18, 1817003(2022)

Adaptive Feature Recombination Recalibration Algorithm for Hepatic Vascular Segmentation

Yuanlu Li1,2、*, Xiangke Shi1, and Kun Li1
Author Affiliations
  • 1School of Automation, Nanjing University of Information Science & Technology, Nanjing 210044, Jiangsu , China
  • 2Jiangsu Atmospheric Environment and Equipment Technology Collaborative Innovation Center, Nanjing 210044, Jiangsu , China
  • show less
    Figures & Tables(11)
    Overall structure of the improved model
    RR block. (a) Overall structure, the residual block is nested with the recombination block, and the recombination block is nested within the SegSE block; (b) structure of the recombination block; (c) structure of the recalibration block; (d) structure of the residual block
    Module structure. (a) SE module; (b) recalibration module
    Attention mechanism structure
    Flow chart of pretreatment
    Broken line diagram of the influence of RR block on the model. (a) Dice Score; (b) Sensitivity
    Segmentation results. (a) CT original image; (b) gold standard; (c) prediction result; (d) comparison between gold standard and prediction result
    Three-dimensional diagram of segmentation results, the left side of each image is the prediction result, and the right side is the comparison between the segmentation result and the real annotation. (a) MPUNet; (b) nnU-Net; (c) Spider UNet; (d) proposed method
    • Table 1. Quantitative analysis of the influence of the RR block on the model

      View table

      Table 1. Quantitative analysis of the influence of the RR block on the model

      ModuleChannelDice /%Sen /%Spe /%Acc /%
      +RR block[48,64]60.8367.9797.9298.34
      [48,64,96]63.9569.7298.9399.12
      [48,64,96,144]61.4371.3298.8598.91
      [48,64,96,144,192]55.9767.9297.7397.87
      -RR block[48,96]30.1333.1193.2493.63
      [48,96,128]37.2148.6594.4395.12
      [48,96,128,256]41.2251.2395.5396.14
      [48,96,128,256,512]33.8541.9194.2194.95
    • Table 2. Influence of each block and loss function on the model

      View table

      Table 2. Influence of each block and loss function on the model

      LossAttentionRR blockDice /%Sen /%Spe /%Acc /%
      Tversky Loss41.2251.2395.5396.14
      +42.2152.3396.1197.34
      +63.9569.7298.9399.12
      ++64.8070.5199.1299.23
      GD Loss40.2252.3195.2296.81
      +41.9553.6396.6197.53
      +62.5371.8198.6098.82
      ++63.7273.1599.3199.10
    • Table 3. Segmentation effect of different models

      View table

      Table 3. Segmentation effect of different models

      ModelYearDice Score
      MPUNet19201959.00
      UMCT20202063.00
      nnU-Net21201963.00
      Spider UNet22202145.60
      C2FNAS-Panc23202064.30
      Proposed model64.80
    Tools

    Get Citation

    Copy Citation Text

    Yuanlu Li, Xiangke Shi, Kun Li. Adaptive Feature Recombination Recalibration Algorithm for Hepatic Vascular Segmentation[J]. Laser & Optoelectronics Progress, 2022, 59(18): 1817003

    Download Citation

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

    Category: Medical Optics and Biotechnology

    Received: Jun. 21, 2021

    Accepted: Aug. 10, 2021

    Published Online: Aug. 29, 2022

    The Author Email: Li Yuanlu (lyl_nuist@nuist.edu.cn)

    DOI:10.3788/LOP202259.1817003

    Topics