Retrospective Study
Copyright ©The Author(s) 2022.
World J Gastroenterol. Nov 7, 2022; 28(41): 5931-5943
Published online Nov 7, 2022. doi: 10.3748/wjg.v28.i41.5931
Table 1 Comparison between the single and ensemble models on GZ_Capcam (artifact-affected) datasets
Model
SegNet
U-Net
Attention-UNet
ResNet-UNet
HarDMSEG
MetricIoUDiceIoUDiceIoUDiceIoUDiceIoUDice
Snap_max0.3410.4540.3790.5070.3610.5020.4140.5220.5380.649
Ens_max0.3410.4620.4060.5320.3750.5090.4350.5370.5500.653
Improve0.0000.0080.0270.0250.0140.0070.0210.0150.0120.004
Improve (%)0.081.717.014.933.881.405.132.862.220.61
Table 2 Comparison between the single and ensemble models on CVC_Colon (clear) datasets
ModelSegNet
UNet
Attention-UNet
ResNet-UNet
HarDMSEG
IoU
Dice
IoU
Dice
IoU
Dice
IoU
Dice
IoU
Dice
Snap_max0.7000.7800.7130.7880.7540.8300.7470.8190.8400.901
Ens_max0.7020.7790.7110.7830.7520.8290.7500.8160.8400.901
Improve0.002-0.001-0.002-0.005-0.002-0.0020.003-0.0020.0000.000
Improve (%)0.30-0.08-0.28-0.61-0.22-0.220.42-0.260.03-0.01
Table 3 Comparison between the single and ensemble models on CVC_Clinic (clear) datasets
ModelSegNet
UNet
Attention-UNet
ResNet-UNet
HarDMSEG
IoU
Dice
IoU
Dice
IoU
Dice
IoU
Dice
IoU
Dice
Snap_max0.8140.8900.8160.8840.8360.9000.8230.8840.8450.893
Ens_max0.8150.8900.8260.8910.8380.8980.8240.8840.8440.892
Improve0.0010.0010.0100.0070.002-0.0010.0010.000-0.001-0.001
Improve (%)0.130.071.200.760.23-0.140.080.02-0.08-0.06