T2-mapping of the sacroiliac joints at 1.5 Tesla. A feasibility and reproducibility study

To evaluate the reproducibility of T2 relaxation time measurements of the sacroiliac joints at 1.5 T. Healthy volunteers underwent an oblique axial multislice multiecho spin-echo sequence of the sacroiliac joints at 1.5 T. Regions of interest were manually drawn using a dedicated software by two musculoskeletal radiologists to include the cartilaginous part of the sacroiliac joints. A senior radiologist performed the measurement twice, while a resident measured once. Intra- and inter-observer reproducibility was tested using the Bland-Altman method. Association between sex and T2 relaxation times was tested using the Mann-Whitney U test. Correlation between T2 relaxation times and body mass index (BMI) was tested using the Spearman's rho. Eighty sacroiliac joints of 40 subjects (mean age: 28 ± 4.8 years, range: 20-43; mean BMI: 23.3 ± 3.1, range: 18.9-30) were imaged. The mean T2 values obtained by the senior radiologist in the first series of measurements were 42 ± 4.4 ms, whereas in the second series were 40.7 ± 4.5 ms. The mean T2 values obtained by the radiology resident were 41.1 ± 4.2 ms. Intra-observer reproducibility was 88% (coefficient of repeatability = 3.8; bias = 1.28; p <.001), while inter-observer reproducibility was 86% (4.7; -.88; p <.001). There was significant association between sex and T2 relaxation times (p =.024) and significant inverse correlation between T2 relaxation times and BMI (r = -.340, p =.002). The assessment of T2 relaxation time measurements of sacroiliac joints seems to be highly reproducible at 1.5 T. Further studies could investigate the potential clinical application of this tool in the sacroiliac joints. (orig.)