Laser & Optoelectronics Progress, Volume. 59, Issue 24, 2415001(2022)

Location Decision of Needle Entry Point Based on Improved Pruning Algorithm

Xiaonan Gao1, Guangyuan Zhang1、*, Fengyü Zhou2, and Dexin Yu3
Author Affiliations
  • 1School of Information Science and Electrical Engineering, Shan Dong Jiao Tong University, Jinan 250375, Shandong, China
  • 2School of Control Science and Engineering, Shandong University, Jinan 250000, Shandong, China
  • 3Department of Radiology, Qilu Hospital of Shandong University, Jinan 250000, Shandong, China
  • show less
    Figures & Tables(28)
    Vein detection algorithm flow of semi-automatic blood collection
    Algorithm flow of automatic detection and annotation of dorsal hand vein injection image
    Structure of AT-U-NET model
    Non-Local structure
    Structure of U-Netup module
    Strengthened feature extraction network
    Original map and segmentation map of dorsal hand vein. (a) Original map; (b) segmentation map
    PT-Pruning flowchart
    Cannibalization stage
    Vein segmentation figure and main line of vascular skeleton. (a) Segmentation; (b) main line
    Decision experiment of needle entry point position
    Dorsal hand vein imaging acquisition equipment
    Original dorsal hand vein images
    Original pictures and label images
    Original pictures and label images
    Detection and segmentation effect of at-u-net dorsal hand vein
    Original back of hand
    Image after homomorphic filtering
    Image after CLAHE processing
    Image after adaptive threshold segmentation
    Image after morphological processing
    Image after closed operation
    Different semantic segmentation model processing effects
    Effect of PT-Pruning needle entry point position decision
    Optimal needle entry point position decision
    • Table 1. Performance indexes of different semantic segmentation models for segmentation of dorsal hand vein

      View table

      Table 1. Performance indexes of different semantic segmentation models for segmentation of dorsal hand vein

      ModelMIOUMPAF1-score
      PSPNet61.4968.3569.8
      U-Net67.9773.8477.6
      SegNet67.3271.5375.1
      RefineNet66.6370.6174.9
      DeepLabv366.4169.4672.8
      AT-U-Net79.5284.9893.6
    • Table 2. Accuracy of needle entry point recognition in effective area

      View table

      Table 2. Accuracy of needle entry point recognition in effective area

      MethodDetection success timesDetection failure timesp1 /%
      Automatic labeling and recognition of dorsal hand vein image238715993.75
      PT-Pruning24638396.73
    • Table 3. Recognition accuracy of optimal needle entry point

      View table

      Table 3. Recognition accuracy of optimal needle entry point

      MethodDetection success timesDetection failure timesp2
      PT-Pruning24578996.5%
    Tools

    Get Citation

    Copy Citation Text

    Xiaonan Gao, Guangyuan Zhang, Fengyü Zhou, Dexin Yu. Location Decision of Needle Entry Point Based on Improved Pruning Algorithm[J]. Laser & Optoelectronics Progress, 2022, 59(24): 2415001

    Download Citation

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

    Category: Machine Vision

    Received: Aug. 31, 2021

    Accepted: Oct. 27, 2021

    Published Online: Jan. 6, 2023

    The Author Email: Zhang Guangyuan (sdjtzhangguangyuan@163.com)

    DOI:10.3788/LOP202259.2415001

    Topics