Table 1.

Evaluation results of nerve segmentation with four learning schemes.

Evaluation Metric	Learning Scheme				p Value
	I	II	III	III’	I vs. II	I vs. III	II vs. III	III vs. III’
sensitivity	0.689 ± 0.391	0.949 ± 0.056	0.962 ± 0.054	0.977 ± 0.025	<0.01	<0.01	0.14	0.16
specificity	0.843 ± 0.147	0.930 ± 0.046	0.937 ± 0.054	0.947 ± 0.030	<0.01	<0.01	0.33	0.19
precision	0.469 ± 0.239	0.719 ± 0.123	0.752 ± 0.118	0.772 ± 0.061	<0.01	<0.01	0.06	0.13
mean accuracy	0.766 ± 0.156	0.939 ± 0.026	0.950 ± 0.031	0.962 ± 0.014	<0.01	<0.01	0.03	0.06
$F_1$ value	0.469 ± 0.243	0.809 ± 0.083	0.837 ± 0.085	0.860 ± 0.034	<0.01	<0.01	0.03	0.05