Performance of a fast and high‐resolution multi‐echo spin‐echo sequence for prostate T2 mapping across multiple systems
Until recently, the application of T2 mapping has been limited by the long acquisition time necessary for accurate measurement of T2 values. This usually resulted in limited spatial coverage and resolution. In the last years, T2 acquisition schemes have been improved, making T2 mapping feasible in clinical practice 6. For example, the standard multi‐echo spin echo (ME‐SE) sequence can be accelerated with a reduced field‐of‐view method using tilted angles 6 or a k‐t undersampling scheme in which acceleration is obtained by skipping lines in k‐space interleaved across multiple echoes 13. Previous studies have shown that the latter yields accurate and artifact‐free prostate T2 maps with voxel sizes close to the standard T2‐weighted images 11.
A key requirement for a T2 mapping sequence to be applicable in a longitudinal or multicenter setting is reliable performance across multiple visits and systems. For the k‐t T2 technique, the repeatability in patients was found to be a few percent by acquiring two T2 maps within the same MRI exam on a single system 11. Bias, short‐term and long‐term repeatability, as well as the reproducibility across multiple were not investigated. Therefore, the purpose of this study is to assess the performance of k‐t T2 maps in prostate cancer. An important aspect of repeatability is its close relation to the size of the regions of interest (ROIs) that are being investigated. Typically, repeatability increases for increasing ROI size, because averaging values over a larger region reduces the effect of image noise 15. Therefore, it is important to estimate repeatability in relation to the smallest ROI size for which changes in T2 values can be detected with the desired reliability. We investigated these aspects with multisystem phantom experiments as well as test‐retest data of prostate cancer patients on a single system.