• Acta Photonica Sinica
  • Vol. 51, Issue 3, 0310001 (2022)
Hong HUANG1、*, Tao WANG1, Yuan LI1, Fanlin ZHOU2, and Yu LI2
Author Affiliations
  • 1Key Laboratory of Optoelectronic Technique System of the Ministry of Education,Chongqing University,Chongqing 400044,China
  • 2Department of Pathology,Chongqing University Cancer Hospital & Chongqing Cancer Institute & Chongqing Cancer Hospital,Chongqing 400030,China
  • show less
    DOI: 10.3788/gzxb20225103.0310001 Cite this Article
    Hong HUANG, Tao WANG, Yuan LI, Fanlin ZHOU, Yu LI. Cancer Pathological Segmentation Network Based on Depth Feature Fusion[J]. Acta Photonica Sinica, 2022, 51(3): 0310001 Copy Citation Text show less
    The overall structure of the proposed HU-Net algorithm
    Fig. 1. The overall structure of the proposed HU-Net algorithm
    The structure of EfficientNet-B4
    Fig. 2. The structure of EfficientNet-B4
    The method of using attention block
    Fig. 3. The method of using attention block
    The structure of attention block
    Fig. 4. The structure of attention block
    The SEED and BOT data sets used in the experiment
    Fig. 5. The SEED and BOT data sets used in the experiment
    Segmentation masks of different algorithms on BOT dataset
    Fig. 6. Segmentation masks of different algorithms on BOT dataset
    Segmentation masks of different algorithms on SEED dataset
    Fig. 7. Segmentation masks of different algorithms on SEED dataset
    AlgorithmDICE /%MA/%Sen/%Pre/%IOU/%
    U-Net70.0686.0468.5777.9358.74
    FCN-VGG1668.0185.0566.7273.6756.91
    SegNet72.0585.8670.8175.5660.30
    DeepLabv3+76.5287.8775.9878.8165.34
    TransUnet75.0287.4978.4075.1465.09
    CA-Net74.3383.4777.5372.1262.00
    HU-Net77.9988.5279.2278.7867.01
    Table 1. The result of different algorithms on the BOT dataset
    AlgorithmDICE/%MA/%Sen/%Pre/%IOU/%
    EU-Net75.2687.0077.5974.5963.73
    FU-Net77.4588.5277.4678.7866.53
    AU-Net76.9887.6476.7878.3665.56
    HU-Net77.9988.6579.2278.9567.01
    Table 2. The results of different models on the BOT dataset
    AlgorithmDice/%MA/%Sen/%Pre/%IOU/%
    U-Net66.3678.9065.6170.1553.65
    FCN-VGG1671.4775.3778.5271.3656.87
    SegNet75.0180.2378.8273.7561.55
    DeepLabv3+81.8686.2482.5980.6669.63
    TransUnet80.2585.1682.9579.7968.94
    CA-Net76.7481.1380.6874.3563.70
    HU-Net82.9487.4284.0182.5672.08
    Table 3. The result of different algorithms on the SEED dataset
    AlgorithmDice/%MA/%Sen/%Pre/%IOU/%
    EU-Net78.9583.4280.8577.5666.10
    AU-Net79.7685.0280.2679.8767.74
    FU-Net81.7686.8482.2381.9270.48
    HU-Net82.9487.4284.082.5672.08
    Table 4. The results of different models on the SEED dataset
    Hong HUANG, Tao WANG, Yuan LI, Fanlin ZHOU, Yu LI. Cancer Pathological Segmentation Network Based on Depth Feature Fusion[J]. Acta Photonica Sinica, 2022, 51(3): 0310001
    Download Citation