Comparison of High-resolution MRI and Optical Microscopy in Quantitation of Trabecular Architecture in the Rat Femur

These calculations were completed for both unfiltered and filtered images. The filtering method removed small objects with a perimeter less than 0.1 mm that did not appear to be trabecular bone. The filtering was mostly only necessary for the optical images due to their very high resolution. In the filtered images the average difference between the MR and optical images for ratio of bone to total volume (BV/TV) was 1.2 %. Correlation between the two methods was lowest for parameters sensitive to trabecular architecture, ranging from r = 0.77 (BV/TV) to r = 0.55 for trabecular number (Tb.N). These relatively low correlations are thought to be caused primarily by distortion of the bone structure in the decalcification process used to section the bone and hence obtain the optical images. MRI of decalcified bone samples prior to sectioning confirmed this. Discussion There is growing evidence to suggest that variations in trabecular structure are important in determining bone strength, independent of bone mineral density (BMD). This work demonstrates the potential of MRI in determining quantitative information on bone morphological parameters. The high resolution achieved in this study was essential to reducing the large partial volume effects normally associated with small animal bone studies. Due to the much larger trabecular architecture in humans a much lower resolution can be used. The ultimate aim of this work is to be able to assess the efficacy of various therapeutic regimes on patients with osteoporosis.