T1 and T2 measurements on a 1.5-T commercial MR imager.

In order for relaxation times to be used in clinical diagnosis, the precision of the measurement must be determined. The authors measured T1, T2, and proton density in a phantom and in human volunteers to determine the reproducibility of the method. The coefficient of variance of T1 measurements in the phantom during a 15-month period with two software upgrades was 5%. Variance of T2 measurements with any given software was 4% or less, and overall in the 15-month period, with two software changes, the T2 reproducibility was between 6% and 9%. The reproducibility is sufficiently high that precise clinical measurements of T1, T2, and proton density are feasible.