Chinese Journal of Lasers, Volume. 49, Issue 20, 2007207(2022)

Automatic Detection of Dental Lesions Based on Deep Learning

Feng Liu1, Min Han1、*, Jun Wang1, and Chao Liu2、**
Author Affiliations
  • 1School of Information Science and Engineering, Shandong University, Qingdao 266237, Shandong, China
  • 2Department of Oral and Maxillofacial Surgery, Qilu Hospital of Shandong University, Jinan 250012, Shandong , China
  • show less
    Figures & Tables(10)
    YOLOV5 structure
    CSP structure
    Network feature fusion realized with FPN+PAN
    Comparison of learning rate strategy convergence. (a) Cosine annealing strategy; (b) one-dimensional linear interpolation strategy; (c) step reduction strategy
    Dental X-ray image and two types of lesions
    Training results display. (a) Bounding box loss; (b) classification loss; (c) confidence loss; (d) precision; (e) recall rate
    P-R curve
    Examples, where caries label represents the are is caries and cyst label represents the area is periapical lesion
    Visualization of thermodynamic diagram. (a) Original drawing; (b) thermodynamic diagram
    • Table 1. Evaluation indices comparison of different algorithms

      View table

      Table 1. Evaluation indices comparison of different algorithms

      AlgorithmP /%R /%A /%F1 score /%
      YOLOV598959596
      SSD91656976
      Faster-RCNN66706368
    Tools

    Get Citation

    Copy Citation Text

    Feng Liu, Min Han, Jun Wang, Chao Liu. Automatic Detection of Dental Lesions Based on Deep Learning[J]. Chinese Journal of Lasers, 2022, 49(20): 2007207

    Download Citation

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

    Category: Biomedical Optical Imaging

    Received: May. 19, 2022

    Accepted: Jun. 17, 2022

    Published Online: Aug. 23, 2022

    The Author Email: Min Han (hanmin@sdu.edu.com), Chao Liu (qiluliuchao@sdu.edu.com)

    DOI:10.3788/CJL202249.2007207

    Topics