Introduction

MRI relaxometry (T2) is a great tool in the detection of osteoarthritis (OA) ¹, which is the most common disease of articular cartilage altering molecular composition (glycosaminoglycan, GAG,) and structural integrity (collagen), together with an increase of water-specific across the entire tibial-femoral joint ². With the addition of µCT measurements for bone-density properties, we show that by accurately combining the two quantitative maps we can accurately study the changes in cartilage and underlying bone which is inadequately measured in MRI. Radiographs of joint space have been widely used for the study of OA progression ³, however with no soft tissue understanding of the changes. Correlation between µCT and µMRI could enhance the knowledge of comprehensive properties of the joint that are unreliably undetectable with conventional MRI.

Methods

Four intact normal rabbit knee joints were imaged using a Bruker AVANCE IIIHD system with a 7T/9 cm vertical-bore magnet with a MiniWB57 probe and a 40 mm coil. All MRI experiments used an MSME sequence with a FOV of 30 mm and a matrix size of 256x128 (post-reconstruct to 512x512), which yielded MRI T2 Maps of the intact joint at 58.6 µm/pixel nominal size. TR was 5000 ms, with 10 incremental echo times (TE) and a min TE at 7.78ms, averaged 8 scans and with 20 slices at 1mm thickness with no gap (hence covering the entire knee joint). The intact knee joints were then imaged in µCT by Skyscan1174 (Bruker, Kontich, Belgium) with experimental parameters set at 50 kV, 800 mA, 12 averages, 0.5° rotation step, 180° rotation, 0.2 mm Al filter, and 652 × 512 data matrix, which took approximately 49 minutes to acquire approximately 608 images with a 60.8 µm isotropic voxel size. All CT images were reconstructed using NRecon (Bruker, Kontich, Belgium) with the same parameters and a global intensity threshold was kept constant for all scans to include the full range of attenuation values.

Results

Figure 1 shows the two-dimensional axial (a), sagittal (b), and coronal (c) µCT images in at 60.8µm/pixel isotropic resolution. The red cross-hair lines in the three images signify the location of slice respective to each other. The coronal slice of the MRI proton-density map (Fig.1.d) is shown compared to the coronal µCT images (d). The red arrows signify the joint space between femur and tibia based on the µCT coronal map. The red arrows on the MRI map signify the same distance shown in the µCT image, where the distance is measured to be between the darkest region under the cartilage in the femoral joint to the tibial joint. Figure 2 shows, the T2 map overlaid with the µCT bone density map (green) of the closest slice location. Here we show the underlying bone properties that are undetectable with MRI.

Discussion

The goal of this study was to correlate the high-resolution MRI T2 maps of intact joint with the finding of bone density properties using isotropic high-resolution µCT. We show with the overlaid µCT maps on the quantitative T2 maps of intact rabbit joint, that by studying the bone density changes, we can easily quantify the changes in the cartilage properties over its full thickness. With the aid of 3D rendering, we can create a knee joint contour based on bone and cartilage (and surrounding tissue) properties. This will aid in future studies of rabbit joint osteoarthritis disease progression due to impact damage in the underlying bone.

Conclusion

High-resolution µCT and MRI studies have shown to benefit from a cross correlative study of osteoarthritic tissue 4. Most clinical MRI studies of topographical knee joint spacing lack a true plateau due to the large slice thickness to improve the signal to noise ratio or adding a slice gap that causes the loss of information due to time constraints in an experimental setup. Here we use the highest isotropic resolution µCT of the whole joint to find the more accurate contour of the tibial and femoral joint, on which we implement the T2 map of cartilage and other soft-tissue properties. We have combined two methods to better understand the contour of the healthy rabbit knee joint, so that in the future when we study an OA tissue, we can easily measure changes in the bone and accurately relate the findings to the MRI T2 map of cartilage.

Acknowledgements

The authors thank NIH for the R01 grant (AR069047).