RESUMO
RATIONALE AND OBJECTIVES: Evaluation of the T2 relaxation time of articular cartilage holds great potential for quantitative assessment of internal changes of the cartilage matrix. The purpose of the present study was to assess the validity of multiecho-based cartilage T2 quantitation in a clinical MRI setting at 1.5 T. METHODS: Four multisection multiecho sequence variants dedicated for quantitative T2 mapping of human articular cartilage were implemented on a 1.5 T whole-body imager and tested for accuracy in CuSO4-agarose gel phantoms and human patellar cartilage. Sequence design was varied to minimize errors in T2 quantitation due to stimulated echoes. RESULTS: As compared with single spin-echo experiments, the apparent T2 values calculated from the multiecho sequence variants showed mean deviations ranging from +26% to -32% (phantoms) and from +42% to -18% (cartilage). The patellar cartilage T2 covered a range from about 25 milliseconds to 55 milliseconds, with longer T2 values observed in the more superficial layers. In cartilage, best results were obtained from the sequence design using improved section profiles and a spoiler gradient scheme for suppression of stimulated echoes. CONCLUSIONS: Our results revealed a clear dependence of apparent T2 relaxation times on the pulse sequence design, emphasizing that the "true" T2 is hard to find. In addition, the effect on the apparent T2 values resulting from the specific modification of any sequence variant varied according to the respective tissue's properties. Therefore, the acquisition technique in conjunction with the specific tissue on which T2 mapping is performed need to be reported in detail and should kept consistent to allow large-scale comparisons and monitoring of treatment strategies, e.g., in osteoarthritis.
