NUCLEAR TECHNIQUES, Volume. 48, Issue 3, 030502(2025)

Radionuclide identification method of two-dimensional convolutional neural network based on nuclear pulse peak sequence

Geng LUO1... Rui SHI1,2,*, Shuxin ZENG1, Yanan SHANG1, Guang YANG1, Yixian DING3, Yadong WU1,2, and Xianguo TUO12,** |Show fewer author(s)
Author Affiliations
  • 1School of Computer Science and Engineering, Sichuan University of Science & EngineeringZigong 643000, China
  • 2Perception and Control Key Laboratory of Sichuan Province, Sichuan University of Science & Engineering,, Yibin 644005, China
  • 3Hongya Ecological Environment Bureau, Meishan 620010, China
  • show less
    Figures & Tables(20)
    Diagram of experimental conditions and data acquisition process
    Sequence scatter plot of 137Cs (a), 60Co (b), 155Eu-22Na (c), 137Cs-60Co (d), 60Co-155Eu-22Na (e), 137Cs-155Eu-22Na (f), 137Cs-60Co-155Eu-22Na (g), background (h)
    Schematic diagram of the mapping of nuclear pulse sequence from one-dimensional to two-dimensional (color online)
    Structure diagram of 2D-CNN network model (color online)
    Train 2D-CNN model loss line graphs with different parameter structures, filling method: 'Same' means zero filling, 'Valid' means no filling, 'BN' means batch normalization, 'NoneBN' means no batch normalization (color online)
    Stratified 10-fold cross-validation accuracy curve: 4 categories (color online)
    Confusion matrix (4 nuclide combinations) (a) Sequence length: 400, (b) Sequence length: 900
    Stratified 10-fold cross-validation accuracy curve: 5 radionuclide combinations (color online)
    Confusion matrix (5 categories) (a) Sequence length: 400, (b) Sequence length: 900
    Stratified 10-fold cross-validation accuracy curve: 8 categories (color online)
    Confusion matrix (8 categories) (a) Sequence length: 400, (b) Sequence length: 900
    Energy spectrum (1) ((a) nuc7, nuc2, (b) nuc6, nuc3) and scatter plot (2) ((c) nuc7, nuc2, (d) nuc6, nuc3) when sequence length is 900, detection distance is 40 cm (color online)
    • Table 1. Information about the radioactive sources used in the experiment

      View table
      View in Article

      Table 1. Information about the radioactive sources used in the experiment

      放射源

      Radioactive sources

      活度

      Radioactivity / Bq

      标签

      Label

      137Cs2.940×105nuc0
      60Co9.900×104nuc1
      155Eu-22Na1.120×104nuc2
      137Cs-60Co3.930×105nuc3
      60Co-155Eu-22Na3.052×105nuc4
      137Cs-155Eu-22Na1.102×105nuc5
      137Cs-60Co-155Eu-22Na4.042×105nuc6

      环境本底

      Environmental background

      无 Nonenuc7
    • Table 2. The number of samples divided into each length sequence set (8 refers to 8 detection points)

      View table
      View in Article

      Table 2. The number of samples divided into each length sequence set (8 refers to 8 detection points)

      样本标签

      Label

      核脉冲序列长度划分

      Nuclear pulse sequence length division

      200:600:1009001 2251 600
      nuc08×1 0008×1 0008×8008×600
      nuc18×1 0008×1 0008×8008×600
      nuc28×1 0008×1 0008×8008×600
      nuc38×1 0008×1 0008×8008×600
      nuc48×1 0008×1 0008×8008×600
      nuc58×1 0008×1 0008×8008×600
      nuc68×1 0008×1 0008×8008×600
      nuc78 0007 6005 6004 200
      合计Total64 00063 60050 40037 800
    • Table 3. The relationship between sequence set length and the corresponding real-time arrival time layer values

      View table
      View in Article

      Table 3. The relationship between sequence set length and the corresponding real-time arrival time layer values

      序列长度 Sequence lengthxmax
      200700
      3001 000
      4001 200
      5001 700
      6002 800
      9004 200
      1 2255 500
      1 6007 200
    • Table 4. Confusion matrix

      View table
      View in Article

      Table 4. Confusion matrix

      类别

      Category

      预测值Predicted
      ŷ=cŷc

      真实值

      True

      y=cTPcFNc
      ycFPcTNc
    • Table 5. Mean recognition accuracy for different model parameters and variance between accuracies

      View table
      View in Article

      Table 5. Mean recognition accuracy for different model parameters and variance between accuracies

      模型参数Model parameter10次测试平均精度Average accuracy of 10 tests / %方差Variance
      2×2 + Valid80.583.96
      2×2 + Same83.751.38
      2×2 + Valid + BN82.002.07
    • Table 6. The detailed structure parameters of the model

      View table
      View in Article

      Table 6. The detailed structure parameters of the model

      输入维度

      Input size

      网络层Layer
      CL1PL1CL2PL2CL3PL3Output
      [10,20,2]Dim[10,20,32][10,5,32][9,4,64][64]4/5/8
      K-num3264
      Field[2, 2][2, 2][2, 2]GAP
      Stride111
      [20,15,2]Dim[20,15,32][10,7,32][9,6,64][4,3,64][3,2,128]1284/5/8
      K-num3264128
      Field[2, 2][2, 2][2, 2][2, 2][2, 2]GAP
      Stride11111
      [20,20,2]Dim[20,20,32][10,10,32][9,9,64][4,4,64][3,3,128]1284/5/8
      [20,25,2]Dim[20,25,32][10,12,32][9,11,64][4-5,64][3-4,128]1284/5/8
      [20,30,2]Dim[20,30,32][10,15,32][9,14,64][4, 7,64][3,6,128]1284/5/8
      [30,30,2]Dim[30,30,32][15,15,32][14,14,64][7,7,64][6,6,128]1284/5/8
      [35,35,2]Dim[35,35,32][17,17,32][16,16,64][8,8,64][7,7,128]1284/5/8
      [40,40,2]Dim[40,40,32][20,20,32][19,19,64][9,9,64][8,8,128]1284/5/8
    • Table 7. The mean time (ms) to collect 400 and 900 pulse sequence points at various detection distances for different radiation sources

      View table
      View in Article

      Table 7. The mean time (ms) to collect 400 and 900 pulse sequence points at various detection distances for different radiation sources

      标签

      Labels

      长度

      Length / cm

      510152025303540
      nuc040040.1868.49105.55153.73200.98252.66297.35340.07
      90097.27154.26237.88346.44453.25568.56669.28766.61
      nuc140039.0669.21104.11147.01191.48244.58292.79344.53
      90098.21155.76234.17331.75432.38551.44659.17774.70
      nuc2400449.83610.92714.32777.13806.52843.04858.08869.49
      9001 010.891 378.851 613.961 753.751 818.561 893.111 927.831 951.92
      nuc340030.3339.5957.8778.93103.97135.57168.41207.66
      90095.0998.49129.41177.71234.39305.61380.21468.16
      nuc440047.7168.88112.77144.31202.04233.65290.96342.26
      900106.86155.78254.07325.60439.70525.41655.11772.39
      nuc540047.1459.0292.53125.15176.06216.98259.01311.06
      900107.78132.57208.17281.54382.06488.61584.71700.99
      nuc640029.9138.5855.2979.65105.57133.81169.23201.72
      90094.4498.95124.85179.47237.36301.35382.67456.34
      nuc7400

      925.07

      2 087.88

      900
    • Table 8. Performance comparison of different methods

      View table
      View in Article

      Table 8. Performance comparison of different methods

      方法/数据集

      Methods/DataSet

      样本类别

      Category

      5 cm10 cm15 cm20 cm25 cm30 cm35 cm40 cm

      测试精度

      Test accuacy

      2D-CNN/序列

      2D-CNN/Sequence

      499.99%100.00%99.95%100.00%99.85%100.00%100.00%99.99%100.00%
      599.92%99.88%99.76%99.56%99.03%96.76%93.76%90.92%98.61%
      899.03%79.22%90.88%78.85%85.47%81.95%76.45%76.45%84.45%

      2D-CNN/能谱

      2D-CNN/Spectrum

      499.92%99.94%100.00%100.00%100.00%100.00%100.00%100.00%99.84%
      599.90%99.98%99.94%99.60%99.16%97.90%96.88%96.10%97.02%
      888.74%81.00%84.14%81.34%81.03%79.25%77.40%73.77%79.37%

      BPNN+PCA/脉冲幅值

      BPNN+PCA/

      Pulse amplitude

      481.35%98.20%99.05%99.15%98.95%98.45%96.40%92.15%95.81%
      570.68%84.80%84.40%85.48%85.00%84.00%84.00%80.56%81.99%
      869.48%69.48%72.26%70.37%73.37%71.34%66.00%65.42%68.83%

      SVM+PCA/脉冲幅值

      SVM+PCA/

      Pulse amplitude

      481.2%99.40%95.90%91.85%85.65%78.55%72.2%66.45%84.13%
      559.28%59.32%50.36%42.84%41.00%41.04%40.76%40.48%47.04%
      854.83%42.57%38.08%31.77%29.37%28.51%28.37%28.10%34.07%
    Tools

    Get Citation

    Copy Citation Text

    Geng LUO, Rui SHI, Shuxin ZENG, Yanan SHANG, Guang YANG, Yixian DING, Yadong WU, Xianguo TUO. Radionuclide identification method of two-dimensional convolutional neural network based on nuclear pulse peak sequence[J]. NUCLEAR TECHNIQUES, 2025, 48(3): 030502

    Download Citation

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

    Category: NUCLEAR PHYSICS, INTERDISCIPLINARY RESEARCH

    Received: Aug. 21, 2024

    Accepted: --

    Published Online: Apr. 15, 2025

    The Author Email: Rui SHI (石睿), Xianguo TUO (庹先国)

    DOI:10.11889/j.0253-3219.2025.hjs.48.240301

    Topics