Quantitative effects of acquisition duration and temporal resolution on the measurement accuracy of prostate dynamic contrast-enhanced MRI data: a phantom study

The aim of this study was to investigate the effect of the temporal resolution (T (res)) and acquisition duration (AD) on the measurement accuracy of contrast concentration-time curves (CTCs), and derived phenomenological and pharmacokinetic parameter values, in a dynamic contrast-enhanced MRI experiment using a novel phantom test device. 'Ground truth' CTCs were established using a highly precise optical imaging system. These precisely known CTCs were produced in an anthropomorphic environment, which mimicked the male pelvic region, and presented to the MRI scanner for measurement. The T (res) was varied in the range [2-24.4 s] and the AD in the range [30-600 s], and the effects on the measurement accuracy were quantified. For wash-in parameter measurements, large underestimation errors (up to 40%) were found using T (res) values >= 16.3 s; however, the measured wash-out rate did not vary greatly across all T (res) values tested. Errors in derived K (trans) and v (e) values were below 14 and 12% for acquisitions with {T (res) <= 8.1 s, AD >= 360 s} and {T (res) <= 16.3 s, AD >= 360 s}, respectively, but increased dramatically outside these ranges. Errors in measured wash-in, wash-out, K (trans), and v (e) parameters were minimised using T (res) <= 8.1 s and AD >= 360 s, with large errors recorded outside of this range.