1 paper2 findings
Image-only model
upin silico1
image-only model predicts lumbar-spine DXA-defined osteoporotic-level bone mineral density in silico (AUROC 0.931 (95% CI 0.912–0.950), versus clinical-only AUROC 0.780 (0.732–0.827)).
“the image-only model achieved AUROC 0.931 (95% CI 0.912-0.950) and the clinical-only baseline 0.780 (0.732-0.827)”
In silicoindependent dual-center validation; DXA T-score ≤ -2.5image-only ablation modellumbar spinen = 3,007
Chest X-rays help flag very low bone density in the lower spine
Osteoporosis and sarcopenia26 Jun 2026
Federated learning-based AI
upin silico1
federated learning-based AI predicts lumbar-spine DXA-defined osteoporotic-level bone mineral density in silico (AUROC 0.942 (95% CI 0.923–0.962); sensitivity 91.0%; specificity 82.7%).
“The AI achieved an AUROC of 0.942 (95% CI, 0.923-0.962), with sensitivity of 91.0% and specificity of 82.7%”
In silicoindependent dual-center validation; DXA T-score ≤ -2.5lumbar spinen = 3,007
Chest X-rays help flag very low bone density in the lower spine
Osteoporosis and sarcopenia26 Jun 2026
Among 3,007 adults, artificial intelligence identified 91.0% of those with very low lower-spine bone density and correctly classified 82.7% of those without it.Among 3,007 adults, artificial intelligence identified 91.0% of those with very low lower-spine bone density and correctly classified 82.7% of those without it.