Skip to main content
. 2021 May 7;21(9):3232. doi: 10.3390/s21093232

Table 2.

Performances of the combination kernel sizes in the two pooling stages by 5-fold cross-validation in MRBrain13 training dataset (DC:%, HD:mm, AVD:%). The “FP” represents the first pooling stage, the “SP” represents the second pooling stage, and the “K” represents the combination of kernels.

GM WM CSF
K K DC HD AVD DC HD AVD DC HD AVD
FP 7,5,3,2 SP 7,5,3,2 82.47 2.10 7.99 83.59 3.61 7.71 78.22 3.30 8.61
FP 5,3,2 SP 5,3,2 83.12 1.94 7.79 85.40 2.89 7.52 81.45 2.58 8.59
FP 5,3,2 SP 3,2 86.08 1.71 6.76 89.02 1.76 6.71 84.15 2.24 7.82
FP 5,3,2 SP 2 84.50 1.75 7.01 86.04 2.75 7.17 83.23 2.44 8.02
FP 3,2 SP 3,2 85.98 1.90 7.22 88.90 2.32 6.59 84.63 2.18 8.13
FP 3,2 SP 2 82.25 2.03 8.07 84.34 3.48 7.42 83.01 2.98 8.66
FP 2 SP 2 85.94 1.85 7.09 88.83 2.39 6.82 83.79 2.31 8.30