Janet Blumenfeld1, Julio Carballido-Gamio2, Jenny Folkesson2, Dana Carpenter2, Thomas M. Link2, Sharmila Majumdar1,2
1Bioengineering, UC-Berkeley/UC-San Francisco, San Francisco, CA, USA; 2Radiology, UC-San Francisco, San Francisco, CA, USA
This study is the first time that image registration has been implemented in a longitudinal study investigating changes in MR-derived trabecular bone structure. Image registration ensures that the same analysis region is used in baseline and follow-up images. The registration technique uses a mutual information based method to accurately register longitudinal MR images of the proximal femur. Results demonstrate improved accuracy in trabecular bone measurements between baseline and one year follow-up images and suggest that this algorithm is robust enough to be used routinely in longitudinal studies of trabecular bone in other musculoskeletal sites, such as the tibia and wrist.