Table 4:

nCBF low- vs high-grade thresholds and ROC analysis results

Team/Entry No.	Software	LG vs HG (P Value)	Threshold	SN	SP	AUC
A5	IB Neuro	2.90E-05	1.815	0.806	0.846	0.861
B5	IB Neuro	2.38E-05	1.555	0.917	0.769	0.864
C2	IB Neuro	3.82E-06	1.415	0.944	0.769	0.880
B6	Matlab	2.61E-04	2.045	0.722	0.846	0.814
D4	nordicICE	9.81E-06	1.255	0.972	0.769	0.868
D5	nordicICE	3.03E-05	1.580	0.861	0.846	0.866
E7	PGUI	1.30E-04	1.575	0.806	0.846	0.846
E8	PGUI	3.85E-05	1.375	0.917	0.692	0.845
E9	PGUI	1.67E-04	1.460	0.833	0.769	0.812
E10	PGUI	7.86E-03	1.920	0.639	0.692	0.658
D6	Slicer	1.04E-06	1.405	0.944	0.769	0.887
F2	FIAT	2.65E-04	1.460	0.861	0.769	0.816

Note:—AUC indicates area under the curve; LG, low-grade; HG, high-grade.