Uncertainty contributions due to	Evaluation method	Relative standard uncert. × 104		Comments
		18F	64Cu
Counting statistics	A	9	12	Counting statistics: Typical standard deviation of the mean for repeated activity determinations (N = 7 or 57 for 18F; N = 45 to 102 for 64Cu) on a single source on a single measurement run (averaged over 2 sources for 18F; average of 4 runs with 3 sources; one source measured on two occasions, with recovered activity consistent to 0.03 % for 64Cu)
Between source variance	A	0.3	34	Standard deviation on the activities determined for multiple sources
Model uncertainty	B	28	32	Estimated by combining the typical difference on the activities recovered with a linear and a quadratic extrapolation (0.27 % for 18F; 0.32 % for 64Cu) with the standard deviation of intercept values obtained with different gamma gates for 18F or the standard deviation on the activity recovered from a linear extrapolation using domains of N = 7 to 12 efficiency points (0.06 %) for 64Cu.
Mass determinations	B	5	5
Live time	B	10	10	Estimated based on previous work
Background	B	3	10	Estimated by propagating the standard deviation of the mean for repeated (N = 10) measurements of the matched blank
Impurities	B	0	6	Based on HPGe measurements and estimates for LS efficiency
Half-life	B	8	2	From DDEP, T1/2(18F) = 1.82890(23) h; T1/2(64Cu) = 12.7004(20) h
β+ branching ratio	B	20		From DDEP, 0.9686(19)
Relative combined standard uncertainty		38	51