Observer concordance and 2 × 2 table values for Extended ASL cohort with both 3.0 T and 1.5 T systems. Observer concordance is defined as in table 1. The left 2 × 2 table is for the entire extended cohort of ASL studies, while the right table is censored, excluding 5 study results that were ambiguous, i.e., outcome was not be definitive by the study's end.

Table 4: Observer concordance and 2 × 2 table values for Extended ASL cohort with both 3.0 T and 1.5 T systems. Observer concordance is defined as in table 1. The left 2 × 2 table is for the entire extended cohort of ASL studies, while the right table is censored, excluding 5 study results that were ambiguous, i.e., outcome was not be definitive by the study's end.

Observer concordance
	Overall agreement	Fleiss' free margin kappa	Complete agreement	Partial discordance (1 v. 2 or 2 v. 3)	Complete discordance (1 v. 3)
ASL total	76%	0.69	52/75 69%	23/75 31%	0/70 0%
ASL censored	80%	0.7	49/70 70%	21/70 30%	0/71 0%

ASL extended cohort
	Cond. +	Cond. -	Sum	PPV, NPV
Test +	14	4	18	0.78
Test -	8	49	57	0.86
Sum	22	53	75
Sens., Spec.	0.64	0.92		0.84

ASL extended cohort censored
	Cond. +	Cond. -	Sum	PPV, NPV
Test +	14	2	16	0.88
Test -	8	46	54	0.85
Sum	22	48	70
Sens., Spec.	0.64	0.96		0.86