Table 3.

Mean values of pairwise comparison metrics for different segmentation models.

MODEL	SPECIFICITY	SENSITIVITY	DICE	95% HD	COHEN’S KAPPA
OURS	0.993	0.826	0.832	1.42 mm	0.823
DENSENET264	0.983	0.847	0.780	1.53 mm	0.772
EFFICIENTNETB5	0.982	0.848	0.767	1.54 mm	0.756
HRNET	0.983	0.843	0.774	1.52 mm	0.763
INCEPTIONRESNETV2	0.981	0.841	0.759	1.57 mm	0.740
INCEPTIONV4	0.981	0.817	0.755	1.54 mm	0.732
RESNEST269	0.979	0.842	0.738	1.61 mm	0.725
RESNET152	0.977	0.848	0.747	1.59 mm	0.710
RESNEXT152	0.981	0.840	0.762	1.55 mm	0.740
SALIENCY GUIDED	0.747	0.900	0.659	1.89 mm	0.516