Voxel size dependency, reproducibility and sensitivity of an in vivo bone loading estimation algorithm

Area of Science:

• Biomechanics
• Medical Imaging
• Computational Biology

Background:

• A previously developed algorithm estimates in vivo bone loading conditions for bone remodeling simulations.
• This algorithm uses a reverse engineering approach, deriving loading history from bone microstructure assessed via high-resolution computed tomography (HR-CT).
• Previous validation showed accurate and realistic results using micro-CT and HR-peripheral quantitative CT (HR-pQCT) images.

Purpose of the Study:

• To investigate the voxel size dependency, reproducibility, and sensitivity of the in vivo bone loading estimation algorithm.
• To determine the optimal voxel size for accurate and reliable bone loading estimations.
• To assess the algorithm's ability to differentiate loading histories based on bone characteristics.

Main Methods:

• Voxel size dependency was tested by downscaling micro-CT images of cadaveric distal radii.
• Reproducibility was assessed using repeated in vitro and in vivo HR-pQCT measurements.
• Sensitivity was evaluated by comparing HR-pQCT images from women with and without fractures, and with varying bone volume fractions.

Main Results:

• The algorithm demonstrated voxel size independence with an average error of 8.2% at 61 µm, but dependency increased significantly beyond 82 µm.
• In vitro and in vivo reproducibility were up to 4.5% and 10.2%, respectively, comparable to existing studies.
• The algorithm distinguished subjects with different bone volume fractions but not those with and without fractures, aligning with adaptation to habitual loads rather than fall impacts.

Conclusions:

• The in vivo bone loading estimation algorithm is reproducible, sensitive, and largely voxel size independent up to 82 µm.
• Smaller voxel sizes are advantageous for improved accuracy and sensitivity in bone loading estimations.
• The algorithm's findings support the understanding of bone adaptation to mechanical loading.