The purpose of this study was to use high resolution magnetic resonance imaging (HR-MRI) combined with structure analysis to investigate the trabecular structure of the human proximal femur and to compare this technique with bone mineral density (BMD) using dual energy X-ray absorptiometry (DXA) in the prediction of bone strength in vitro. Thirty-one fresh human proximal femur specimens were examined with HR-MRI using a T1-weighted 3D spinecho-sequence in a coronal plane (voxel size: 0.195 x 0.195 x 0.9 mm and 0.195 x 0.195 x 0.3 mm). In these images structure parameters analogous to standard bone histomorphometry were obtained in a femoral head, neck, and trochanteric region of interest (ROI). In addition, BMD measurements were obtained using DXA and finally, all specimens were tested biomechanically in a materials testing machine, and maximum compressive strength (MCS) was determined. Correlations between BMD and MCS were significant (p <0.01) with R-values up to 0.74. Correlating structure parameters and MCS R-values up to 0.69 (P <0.01) were obtained. Using multivariate regression analysis, combining structure parameters and BMD, improved correlations versus MCS substantially (up to R = 0.93; P <0.01). In conclusion, this study showed that in an experimental setting, structure parameters determined in high resolution MR images of the proximal femur correlated significantly with bone strength. The highest correlations, however, were obtained combining BMD and structure measures.

The purpose of this study was to compare trabecular bone structure parameters obtained from high-resolution magnetic resonance (HRMR) and multislice computed tomography (MSCT) images with those determined in contact radiographs from corresponding specimen sections. High-resolution MR and MSCT images were obtained in 39 distal radius specimens. For HRMR the in-plane spatial resolution was 0.152×0.153 mm 2 with a slice thickness of 0.9 and 0.3 mm using a 3D T1-weighted spin-echo sequence. For MSCT the resolution was 0.247×0.247 mm 2 with a collimation of 1 mm. Using a diamond saw, 117 0.9-to 1-mm-thick sections were obtained from these specimens and contact radiographs were acquired. In the corresponding sections structure parameters analogous to bone histomorphometry were determined. Significant correlations between MR-and CT-derived structure pa-rameters and those derived from the contact radiographs were found (p<0.01); r values of up to 0.75 were obtained for HRMR imaging and up to 0.70 for MSCT. On the average, structure parameters showed higher correlations for the MR-than for the CT-derived data. For the MR data the threshold algorithm used for binarizing the images substantially affected these correlations. In conclusion, trabecular bone structure parameters assessed in distal radius HRMR and MSCT images are significantly correlated with those determined in corresponding specimen sections (p<0.01). High-resolution MR-derived structure parameters, however, performed better in the prediction of trabecular bone structure.

No abstract

Trabecular bone structure depicted by HR-MR images is significantly correlated with that shown in macro sections (P < 0.01); however, a number of limitations have to be considered, including the substantial impact of thresholding techniques and slice thickness.

MR imaging-derived structure measurements in the calcaneus are useful for monitoring bone changes after heart transplantation and assessing vertebral fracture status.