Figure 4

Sample size determination using the coefficient of variance for unpaired data in 3D-TOF-MRA studies where day-to-day and side-to-side differences of the middle meningeal (MMA) and middle cerebral (MCA) arterial calibre are desired (5% significance, 90% power). For paired data, the sample size is half of the required sample size should be used: N = 21*(coefficient of variance)² / (lnμ₀ – lnμ₁)².