Table 16.

Compiled results of Fig. 13: Influence of training set size. Comparison of FastSurferVINN trained on the original training set (n = 120) and a version with approximately 20 times more 1.0 mm scans (n = 1315). Dice Similarity Coefficient (DSC) and average surface distance (ASD) are reported for four different resolutions (0.7 mm, 0.8 mm, 0.9 mm and 1.0 mm).

	Datasets	FastSurferVINN + inSA (n = 120)	FastSurferVINN + inSA (n = 1300)		FastSurferVINN + inSA (n = 120)	FastSurferVINN + inSA (n = 1300)
DSC	0.7 mm	89.4 (± 1.29)	89.87 (± 1.28)	ASD	0.259 (± 0.08)	0.243 (± 0.080)
	0.8 mm	89.84 (± 1.46)	90.57 (± 1.44)		0.241 (± 0.079)	0.218 (± 0.080)
	0.9 mm	85.45 (± 1.13)	88.75 (± 1.39)		0.347 (± 0.050)	0.263 (± 0.052)
	1.0 mm	87.67 (± 2.51)	90.26 (± 1.54)		0.296 (± 0.105)	0.231 (± 0.074)