Table 5.

Qualitative results from visual scoring of automatic inferior alveolar nerve segmentation on cone beam computed tomography from 53 randomly selected data (Internal: A, External: B and C)

Grade	Manual						nnU-net (Last step of active learning)
	A		B		C		A		B		C
	Rt	Lt	Rt	Lt	Rt	Lt	Rt	Lt	Rt	Lt	Rt	Lt
4 (very accurate)	14.3	14.3	19	19	19	19	11	9.7	10.7	12.3	5.3	4.7
3 (accurate)	0.7	0.7	0	0	0	0	2.7	3	7.3	5	8.7	7
2 (mostly accurate)	0	0	0	0	0	0	1.3	2	1	1.4	3.3	6.3
1 (inaccurate)	0	0	0	0	0	0	0	0.3	0	0.3	1.7	1

A, Korea University Anam Hospital; B, Korea University Ansan Hospital; C, Korea University Guro Hospital; Rt, right; Lt, left