Table 3:

ROC curve analysis of visual and automated collateral measured for discriminating good tissue outcome (FIV < 90 mL and PPS > 50%)

Collateral Measure	Final Infarct <90 mL						Penumbra Saved >50%
Collateral Measure	AUC (± SE)	P	Optimal Criterion	Youden Index	Sens. (%)	Spec. (%)	AUC (± SE)^{^a}	P	Optimal Criterion	Youden Index	Sens. (%)	Spec. (%)
Visual scoring (CS)
CE-MRA	.81 (.07)	<.01^{^b}	≥1	.53	86.2	66.7	.67 (.10)	.04^{^b}	>1	.27	36.4	90.9
TOF	.53 (.02)	.15	≥1	–	–	–	.53 (.05)	.52	–	–	–	–
TOF + CE-MRA	.81 (.07)	<.01^{^b}	≥1	.53	86.2	66.7	.65 (.10)	.16	–	–	–	–
Atlas-based collateral index (CI)
CE-MRA	.94 (.04)	<.01^{^b}	>.68	.76	89.7	86.7	.83 (.10)	<.01^{^b}	>.64	.70	87.9	81.8
TOF	.74 (.08)	.01^{^b}	>.37	.43	75.9	66.7	.86 (.06)	<.01^{^b}	>.38	.58	75.8	81.8
TOF + CE-MRA	.95 (.03)	<.01^{^b}	>1.13	.80	86.2	93.3	.89 (.08)	<.01^{^b}	>1.05	.76	84.9	9.9

Note:—Sens. indicates sensitivity; Spec., specificity.
↵a AUC indicates area under the ROC curve with associated significance level P. Optimal criterion defines collateral parameter cutoff for maximized sensitivity and specificity by the Youden index.
↵b Significant.