PURPOSE: To analyze the 3D microarchitecture of rat lumbar vertebrae by micro-magnetic resonance imaging (micro-MRI). MATERIALS AND METHODS: micro-MR images (20 x 20 x 20 microm(3) apparent voxel size) were acquired with a three-dimensional spin-echo pulse sequence on four lumbar vertebrae of two rats. Apparent microarchitectural parameters like trabecular bone fraction (BV/TV), specific bone surface (BS/TV), mean intercept length (MIL), and Euler number per unit volume (Euler density, E(V)) were calculated using a novel semiquantitative variable threshold segmentation technique. The threshold value T was obtained as a point of minimum or maximum of the function E(V) = E(V)(T). RESULTS: Quantitative 3D analysis of micro-MRI images revealed a higher connectivity in the peripheral regions (E(V) = -570 +/- 70 mm(-3)) than in the central regions (E(V) = -130 +/- 50 mm(-3)) of the analyzed rat lumbar vertebrae. Smaller intertrabecular cavities and larger bone volume fractions were observed in peripheral regions as compared to central ones (MIL = 0.18 +/- 0.01 mm and 0.26 +/- 0.01 mm; BV/TV = 34 +/- 3% and 29 +/- 3%, respectively). The quantitative 3D study of MIL showed a structural anisotropy of the trabeculae along the longitudinal axis seen on the images. The inhomogeneity of the bone architecture was validated by micro-computed tomography (micro-CT) images at the same spatial resolution. CONCLUSION: 3D high-field micro-MRI is a suitable technique for the assessment of bone quality in experimental animal models. (c) 2009 Wiley-Liss, Inc.
PURPOSE: To analyze the 3D microarchitecture of rat lumbar vertebrae by micro-magnetic resonance imaging (micro-MRI). MATERIALS AND METHODS: micro-MR images (20 x 20 x 20 microm(3) apparent voxel size) were acquired with a three-dimensional spin-echo pulse sequence on four lumbar vertebrae of two rats. Apparent microarchitectural parameters like trabecular bone fraction (BV/TV), specific bone surface (BS/TV), mean intercept length (MIL), and Euler number per unit volume (Euler density, E(V)) were calculated using a novel semiquantitative variable threshold segmentation technique. The threshold value T was obtained as a point of minimum or maximum of the function E(V) = E(V)(T). RESULTS: Quantitative 3D analysis of micro-MRI images revealed a higher connectivity in the peripheral regions (E(V) = -570 +/- 70 mm(-3)) than in the central regions (E(V) = -130 +/- 50 mm(-3)) of the analyzed rat lumbar vertebrae. Smaller intertrabecular cavities and larger bone volume fractions were observed in peripheral regions as compared to central ones (MIL = 0.18 +/- 0.01 mm and 0.26 +/- 0.01 mm; BV/TV = 34 +/- 3% and 29 +/- 3%, respectively). The quantitative 3D study of MIL showed a structural anisotropy of the trabeculae along the longitudinal axis seen on the images. The inhomogeneity of the bone architecture was validated by micro-computed tomography (micro-CT) images at the same spatial resolution. CONCLUSION: 3D high-field micro-MRI is a suitable technique for the assessment of bone quality in experimental animal models. (c) 2009 Wiley-Liss, Inc.