eLife assessment
This study presents a valuable deep learning-based model for predicting fracture within the next five years from just a standard distal radius and ulna scan obtained using high-resolution computed tomography images. The evidence supporting the conclusion that the model-predicted fracture prediction score can be used clinically to identify women at risk of fracture more effectively than with the current standard clinical approach is convincing. This work will be of interest to biomechanists and biomedical engineers working on osteoporosis.