The robustness of T2 value as a trabecular structural index at multiple spatial resolutions of 7 Tesla MRI.

ABSTRACT

PURPOSE: To evaluate the robustness of MR transverse relaxation times of trabecular bone from spin-echo and gradient-echo acquisitions at multiple spatial resolutions of 7 T. METHODS: The effects of MRI resolutions to T2 and T2* of trabecular bone were numerically evaluated by Monte Carlo simulations. T2 , T2*, and trabecular structural indices from multislice multi-echo and UTE acquisitions were measured in defatted human distal femoral condyles on a 7 T scanner. Reference structural indices were extracted from high-resolution microcomputed tomography images. For bovine knee trabecular samples with intact bone marrow, T2 and T2* were measured by degrading spatial resolutions on a 7 T system. RESULTS: In the defatted trabecular experiment, both T2 and T2* values showed strong ( |r| > 0.80) correlations with trabecular spacing and number, at a high spatial resolution of 125 µm3 . The correlations for MR image-segmentation-derived structural indices were significantly degraded ( |r| < 0.50) at spatial resolutions of 250 and 500 µm3 . The correlations for T2* rapidly dropped ( |r| < 0.50) at a spatial resolution of 500 µm3 , whereas those for T2 remained consistently high ( |r| > 0.85). In the bovine trabecular experiments with intact marrow, low-resolution (approximately 1 mm3 , 2 minutes) T2 values did not shorten ( |r| > 0.95 with respect to approximately 0.4 mm3 , 11 minutes) and maintained consistent correlations ( |r| > 0.70) with respect to trabecular spacing (turbo spin echo, 22.5 minutes). CONCLUSION: T2 measurements of trabeculae at 7 T are robust with degrading spatial resolution and may be preferable in assessing trabecular spacing index with reduced scan time, when high-resolution 3D micro-MRI is difficult to obtain.