Automated quantitative analysis of peri-articular bone microarchitecture in HR-pQCT knee images

We have developed and evaluated an automated workflow for peri-articular analysis of knee HR-pQCT images, integrating deep learning, atlas-based segmentation, and standard image processing approaches. The code, atlases, and models have been made freely available for other researchers to use, improve, or extend. Future work will focus on the application of the workflow to clinical data to investigate osteoarthritis etiology.