Table 1 Performance evaluation of kidney tumor segmentation according to labeled data ratios in supervised and semi-supervised learning network

Methods	Network	Ratio(%)		Metrics(%)
Methods	Network	Labeled	Unlabeled	Balanced accuracy	F1-score	Recall	Precision
Supervised learning	U-Net[24]	100	0	69.7±16.0	43.4±33.0	39.7±32.3	61.3±39.2
	U-Net[24]	25	0	64.4±14.3	30.5±28.8	29.5±29.0	39.5±36.3
	nnU-Net[25]	100	0	78.9±17.7	61.2±34.4	58.0±35.6	76.9±34.1
	nnU-Net[25]	25	0	68.7±18.5	41.6±38.6	37.8±37.1	59.7±44.7
Semi supervised learning	MT[23]	25	75	63.4±13.2	26.2±27.2	27.5±26.7	34.6±35.0
Semi supervised learning	MT with Kidney Local Guided Map (Ours)	25	75	86.4±14.7	75.2±26.8	72.9±29.4	86.5±24.6