Table B.8.

Results (mean ± standard deviation) of the multiclass segmentation tasks for MRA and CTA in terms of class average Dice scores, centerline Dice (clDice) scores and class average errors in the zero-th Betti number β₀. The clDice scores are computed on the merged binary class, the Dice scores and β₀ errors are computed for each class separately and the average is taken per case. The arrow indicates the favorable direction. The top three values for each metric and each track are marked as gold, silver and bronze cells in decreasing order. If a team only submitted to one of the tracks the columns of the other track are filled with a ‘-’. The winning team names are in bold.

Multi-Class segmentation performance
	MRA			CTA
Team	Per case class-average Dice (%) ↑	clDice (%) ↑	Per case class-average β0 error ↓	Per case class-average Dice (%) ↑	clDice (%) ↑	Per case class-average β0 error ↓
agaldran	0.01 ± 0.07	0.16 ± 0.69	1.01 ± 0.18	0.15 ± 0.66	0.77 ± 1.62	1.07 ± 0.32
gbCoW	80.51 ± 14.69	98.10 ± 2.33	0.28 ± 0.27	-	-	-
gl	81.27 ± 9.16	95.81 ± 5.11	0.09 ± 0.08	54.02 ± 21.76	75.80 ± 28.05	0.24 ± 0.22
IWantToGoToCanada	-	-	-	74.34 ± 7.90	97.61 ± 1.78	0.77 ± 0.50
junqiangchen	71.56 ± 9.72	96.86 ± 2.80	0.87 ± 0.41	67.68 ± 5.97	95.91 ± 1.63	1.09 ± 0.37
lWM	79.03 ± 8.75	96.23 ± 3.11	0.27 ± 0.22	72.49 ± 8.46	92.78 ± 5.13	0.32 ± 0.19
NexToU	83.76 ± 6.95	97.29 ± 2.29	0.14 ± 0.14	81.67 ± 6.81	97.78 ± 1.88	0.16 ± 0.14
NIC-VICOROB-1	77.13 ± 20.15	96.31 ± 5.78	0.11 ± 0.09	51.84 ± 31.11	97.86 ± 2.38	0.43 ± 0.38
NIC-VICOROB-2	65.99 ± 8.76	94.12 ± 6.12	2.64 ± 3.00	62.41 ± 8.22	95.65 ± 3.38	2.31 ± 1.37
Organizers	83.98 ± 7.33	98.06 ± 2.30	0.16 ± 0.13	77.00 ± 11.95	96.98 ± 3.48	0.23 ± 0.26
refrain	83.72 ± 5.79	98.20 ± 1.84	0.19 ± 0.17	-	-	-
sjtu_eiee_2–426lab	-	-	-	67.46 ± 27.78	97.15 ± 3.22	0.23 ± 0.27
WilliWillsWissen	84.58 ± 6.47	97.21 ± 3.37	0.06 ± 0.06	83.32 ± 5.65	98.12 ± 1.86	0.11 ± 0.11