Acta Photonica Sinica, Volume. 51, Issue 6, 0618002(2022)

A Fluorescence Lifetime Retrieval Algorithm Based on LSTM Neural Network

Biyu YANG1 and Yue XU1,2、*
Author Affiliations
  • 1College of Integrated Circuit Science and Engineering,Nanjing University of Posts and Telecommunications,Nanjing 210023,China
  • 2National and Local Joint Engineering Laboratory of RF Integration and Micro-assembly Technology,Nanjing 210023,China
  • show less
    Figures & Tables(13)
    TCSPC method for fluorescence lifetime detection
    Flow chart of the construction of the LSTM neural network model
    Structure of LSTM
    Test results of the same neurons with the different layers
    Test results of the different neurons with the same layers
    The structure of the network
    Histogram of fluorescence lifetime decay
    The restoration results of 32×32 array
    Comparison of CMM,LSM and LSTM retrieval range
    The comparison of array imaging when τ1=5 ns,τ2=10 ns,τ3=15 ns
    The comparison of array imaging when τ1=40 ns,τ2=60 ns,τ3=80 ns
    • Table 1. Comparison of CMM、LSM and LSTM algorithms for retrieving single point when the photon count is 5 000

      View table
      View in Article

      Table 1. Comparison of CMM、LSM and LSTM algorithms for retrieving single point when the photon count is 5 000

      Theoretical fluorescence lifetime τ/nsRetrieval algorithmRetrieval value/nsRacc/%

      Standard

      deviation/ps

      5CMM5.4690.80%88.40
      LSM5.4790.60%186.20
      LSTM4.9098.00%98.26
      10CMM10.8791.30%158.08
      LSM10.9890.20%355.92
      LSTM10.1298.80%181.25
      40CMM31.3678.40%302.40
      LSM45.2486.90%1 892.79
      LSTM41.4696.32%1 383.94
      80CMM38.8948.61%320.77
      LSM92.1384.84%4 892.03
      LSTM81.6597.94%3 392.69
      90CMM39.7344.14%332.59
      LSM103.6384.86%5 540.91
      LSTM86.8796.52%2 794.33
    • Table 2. Comparison of CMM、LSM and LSTM algorithms for retrieving single point when the photon count is 10 000

      View table
      View in Article

      Table 2. Comparison of CMM、LSM and LSTM algorithms for retrieving single point when the photon count is 10 000

      Theoretical fluorescence lifetime τ/nsRetrieval algorithmRetrieval value/nsRacc/%

      Standard

      deviation/ps

      5CMM5.4591.00%54.20
      LSM5.3992.20%93.78
      LSTM5.0998.20%61.48
      10CMM10.7392.70%122.72
      LSM10.8191.90%231.10
      LSTM9.8998.90%106.12
      40CMM30.3675.90%229.64
      LSM40.9797.58%834.33
      LSTM40.7198.23%831.88
      80CMM39.5749.46%218.26
      LSM90.3687.05%2 495.16
      LSTM80.1899.78%2 263.88
      90CMM40.5645.07%236.90
      LSM101.6287.09%3 088.81
      LSTM87.7297.47%2 301.81
    Tools

    Get Citation

    Copy Citation Text

    Biyu YANG, Yue XU. A Fluorescence Lifetime Retrieval Algorithm Based on LSTM Neural Network[J]. Acta Photonica Sinica, 2022, 51(6): 0618002

    Download Citation

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

    Category: Microscopy

    Received: Jan. 17, 2022

    Accepted: Feb. 22, 2022

    Published Online: Sep. 23, 2022

    The Author Email: XU Yue (yuex@njupt.edu.cn)

    DOI:10.3788/gzxb20225106.0618002

    Topics