Laser & Optoelectronics Progress, Volume. 62, Issue 14, 1417002(2025)

Medical Image Segmentation Method Combining a Pyramid Vision Transformer and a Kolmogorov-Arnold Network

Zhongan Huang1, Xinyu Li2, Qiaohong Liu3, min Lin4、**, and Huayuan Yang1、*
Author Affiliations
  • 1School of Acupuncture-Moxibustion and Tuina, Shanghai University of Traditional Chinese Medicine, Shanghai 201203, China
  • 2School of Health Science and Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China
  • 3College of Medical Instruments, Shanghai University of Medicine and Health Sciences, Shanghai 201318, China
  • 4Shanghai Tongji Hospital, Shanghai 200065, China
  • show less
    Figures & Tables(9)
    PVT-KANet model network structure
    Structure of MSCA Tok-KAN
    KAN layer structure
    MSCA block
    Inception depthwise convolutional decoder block
    Comparison of segmentation results of different models on three datasets
    Qualitative analysis results in complex scenarios
    • Table 1. Comparison results of different models on three datasets

      View table

      Table 1. Comparison results of different models on three datasets

      MethodYearParams /106CVC-ClinicDBBUSIGlaS
      IoUF1IoUF1IoUF1
      U-Net201531.03883.7991.0657.2271.9186.6692.79
      U-Net++201836.63084.6191.5357.4172.1187.0792.96
      U-NeXt20191.47274.8385.3659.0673.0884.5191.55
      Rooling-UNet202428.32182.8790.4861.0074.6786.4292.63
      U-Mamba2024173.53084.7991.6361.8175.5587.0193.02
      U-KAN20249.38585.0591.8863.3876.4087.6493.37
      Proposed method40.44689.7194.5565.0278.4888.2193.73
    • Table 2. The ablation experimental results of the proposed model on three datasets

      View table

      Table 2. The ablation experimental results of the proposed model on three datasets

      ModelCVC-ClinicDBBUSIGlaS
      BaselinePVTMSCAIDCDIoUF1IoUF1IoUF1
      85.5592.1262.9176.5685.4792.15
      88.7293.9963.6477.4987.8293.50
      87.2993.1763.7377.4586.4892.36
      89.7194.5565.0278.4888.2193.73
    Tools

    Get Citation

    Copy Citation Text

    Zhongan Huang, Xinyu Li, Qiaohong Liu, min Lin, Huayuan Yang. Medical Image Segmentation Method Combining a Pyramid Vision Transformer and a Kolmogorov-Arnold Network[J]. Laser & Optoelectronics Progress, 2025, 62(14): 1417002

    Download Citation

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

    Category: Medical Optics and Biotechnology

    Received: Dec. 11, 2024

    Accepted: Feb. 7, 2025

    Published Online: Jul. 16, 2025

    The Author Email: min Lin (linm_doc@163.com), Huayuan Yang (yhyabcd@sina.com)

    DOI:10.3788/LOP242398

    CSTR:32186.14.LOP242398

    Topics