图共 5个 表共 6
    • 图  FIG. 487.  FIG. 487.

      Figure FIG. 487..  FIG. 487.

    • 图  1  Dataset 1(a)和Dataset 2(b)的实验范式

      Figure 1.  Experimental paradigms of dataset 1 (a) and dataset 2 (b)

    • 图  2  MCFS算法导联选择流程图

      Figure 2.  Flow chart of channel selection with MCFS algorithm

    • 图  3  各个受试者在AC和MCFS算法下的导联选择数

      Figure 3.  Numbers of channels selected by each subject under AC and MCFS algorithm

    • 图  4  Dataset 1和Dataset 2的导联选择分布图

      Figure 4.  Distribution of selected channels for dataset 1 and dataset 2

    • SubjectTrust value
      MICSPCA
      a1.9×1091.8×10−92×10−13
      b3.7×10−91.12.5×10−13
      f8.3×10241.1×10−29.9×10−1
      g1.1×10112.7×10101.1×105
      aa4.5×10667.7×10372.3×1032
      al1.4×101571.5×10151.5×1015
      av4×101192.2×1032.2×103
      aw101132.4×10114.1×1015
      ay7.7×101982.2×10204.6×1021

      表 1  各受试者对于3种相关性算法的信任值

      Table 1.  Trust values of three correlation algorithms for each subject

    • MethodDataset 1Datasets 2
      abfgMeanaaalavawayMean
      AC5959595959118118118118118118
      MCFS114132413252852511935

      表 2  各个受试者在AC和MCFS算法下的导联选择数

      Table 2.  Numbers of channels selected by each subject under AC and MCFS algorithm

    • SubjectAccuracy/%
      AC3CCSP-rankMCFS
      a687579.7583
      b49.55653.5563
      f62.567.574.183.5
      g83.5878194.5
      Mean±std65.9±14.171.4±1372.1±12.781±13.1

      表 3  Dataset 1在不同导联选择算法下的分类精度

      Table 3.  Classification accuracy of dataset 1 with different algorithms

    • SubjectAccuracy/%
      AC3CCSP-rankMCFS
      aa62.1466.4372.3680.71
      al9089.2992.5497.14
      av61.4357.8561.6172.86
      aw84.2978.9383.1893.21
      ay76.7989.2984.1192.86
      Mean±std74.9±12.976.4±1478.8±1287.4±10.2

      表 4  Dataset 2在不同导联选择算法下的分类精度

      Table 4.  Classification accuracy of dataset 2 with different algorithms

    • MethodsAccuracy/%
      abfgMean
      Binary-SDMOPSO[15]62.3155.560.076460.47
      Improved SFFS[16]6963657267.25
      CCS-CSP[12]80.50607893.5078
      MCFS836383.5094.5081

      表 5  Dataset 1在不同导联选择算法下的分类精度

      Table 5.  Comparison of classification accuracy with different algorithms for dataset 1

    • MethodsAccuracy/%
      aaalavawayMean
      RD-CSP[17]74.1098.2059.2077.7080.6078
      SCSP[18]80.7197.1457.148591.4282.28
      WOLA-CSP[19]66.7996.0752.1471.435067.29
      CSP-R-MF[20]81.4392.417083.578582.48
      MCFS80.7197.1472.8693.2192.8687.40

      表 6  Dataset 2在不同导联选择算法下的分类精度

      Table 6.  Comparison of classification accuracy with different algorithms for dataset 2