T2 relaxation time in patellar cartilage - Global and regional reproducibility at 1.5 Tesla and 3 Tesla

Purpose: Evaluation of the global and regional reproducibility of T2 relaxation time in patellar cartilage at 1.5 T and 3 T. Materials and Methods: 6 left patellae of 6 healthy volunteers (aged 25-30, 3 female, 3 male) were examined using a fat-saturated multiecho sequence and a T1-w 3D-FLASH sequence with water excitation at 1.5 Tesla and 3 Tesla. Three consecutive data sets were acquired within one MRI session with the examined knee being repositioned in the coil and scanner between each data set. The segmented cartilage (FLASH sequence) was overlaid on the multiecho data and T2 values were calculated for the total cartilage, 3 horizontal layers consisting of a superficial, intermedial and deep layer, 3 facets consisting of a medial, median (ridge) and lateral facet (global T2 values) and 27 ROls/MRI slices (regional T2 value). The reproducibility (precision error) was calculated as the root mean square average of the individual standard deviations [ms] and coefficients of variation (COV) [%]. Results: The mean global reproducibility error for T2 was 3.53% (+/- 0.38%) at 1.5 Tesla and 3.25% (+/- 0.61 %) at 3 Tesla. The mean regional reproducibility error for T2 was 8.62% (+/- 2.61%) at 1.5 Tesla and 9.66% (+/- 3.37%) at 3 Tesla. There was no significant difference with respect to absolute reproducibility errors between 1.5 Tesla and 3 Tesla at a constant spatial resolution. However, different reproducibility errors were found between the cartilage layers. One third of the data variability could be attributed to the influence of the different cartilage layers, and another 10% to the influence of the separate MRI slices. Conclusion: Our data provides an estimation of the global and regional reproducibility errors of T2 in healthy cartilage. In the analysis of small subregions, an increase in the regional reproducibility error must be accepted. The data may serve as a basis for sample size calculations of study populations and may contribute to the decision regarding the level of detail of an evaluation of study data.