Artificial Scientific Intelligence for Measurement-burden-aware Modelling and Interpretation of Multi-site Bone Mineral Density
- 1Nation University of Singapore
Abstract
Agentic workflows can coordinate modelling, but balancing predictive performance, measurement burden and reproducibility is unclear. We developed DXA Agent, an agentic workflow for dual-energy X-ray absorptiometry (DXA) outcomes integrating planning, feature-model refinement, tools, provenance and hypothesis-generating interpretation. Models were independently developed and tested in UK Biobank (5,318 participants) and the National Health and Nutrition Examination Survey (NHANES; 3,777 participants), using cost-efficient and no-limit strategies. Across 20 UK Biobank and three NHANES bone mineral density sites, cost-efficient models achieved lower RMSE and higher R{superscript 2} than the best conventional comparator, with median relative RMSE reductions of 10.9% and 9.9%, respectively. Classification was task dependent: UK Biobank osteoporosis averaged AUROC 0.839 and PR-AUC 0.182, whereas NHANES performance was comparable with conventional models. Higher-burden features did not consistently improve prediction. These retrospective, cohort-internal findings position DXA Agent as an inspectable, measurement-burden-aware research workflow requiring independent prospective validation.
Related
- Artificial Intelligence Echocardiographic Interpretation in the Emergency Department: Agreement, Interpretation Yield and Reporting Scope
- CT-Referenced Diagnostic Performance and Complementary Error Patterns of Artificial Intelligence Versus a Single Emergency Physician in Chest Radiograph Interpretation: A Retrospective Paired Diagnostic Accuracy Study
- Validating Artificial Intelligence Guidance for Ultrasound Acquisition and Remote Interpretation
- Artificial Intelligence Versus Human Interpretation in Diagnostic Radiology: A Systematic Review of Diagnostic Accuracy and Operational Efficiency.
- Generative artificial intelligence in the preparation of scientific medical articles on surgery: principles of responsible use and current requirements of editorial organisations
BibTeX
@misc{xiang2026artificial,
title = {{Artificial Scientific Intelligence for Measurement-burden-aware Modelling and Interpretation of Multi-site Bone Mineral Density}},
author = {S. Xiang and H. He and Z. Xie and C.-Y. Cheng and H. Li and D. Liu},
howpublished = {medRxiv},
year = {2026},
month = aug,
doi = {10.64898/2026.08.30.26361665},
url = {https://www.medrxiv.org/content/10.64898/2026.08.30.26361665v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Artificial Scientific Intelligence for Measurement-burden-aware Modelling and Interpretation of Multi-site Bone Mineral Density
AU - S. Xiang
AU - H. He
AU - Z. Xie
AU - C.-Y. Cheng
AU - H. Li
AU - D. Liu
PY - 2026
DA - 2026/08/30
PB - medRxiv
DO - 10.64898/2026.08.30.26361665
UR - https://www.medrxiv.org/content/10.64898/2026.08.30.26361665v1
AB - Agentic workflows can coordinate modelling, but balancing predictive performance, measurement burden and reproducibility is unclear. We developed DXA Agent, an agentic workflow for dual-energy X-ray absorptiometry (DXA) outcomes integrating planning, feature-model refinement, tools, provenance and hypothesis-generating interpretation. Models were independently developed and tested in UK Biobank (5,318 participants) and the National Health and Nutrition Examination Survey (NHANES; 3,777 participants), using cost-efficient and no-limit strategies. Across 20 UK Biobank and three NHANES bone mineral density sites, cost-efficient models achieved lower RMSE and higher R{superscript 2} than the best conventional comparator, with median relative RMSE reductions of 10.9% and 9.9%, respectively. Classification was task dependent: UK Biobank osteoporosis averaged AUROC 0.839 and PR-AUC 0.182, whereas NHANES performance was comparable with conventional models. Higher-burden features did not consistently improve prediction. These retrospective, cohort-internal findings position DXA Agent as an inspectable, measurement-burden-aware research workflow requiring independent prospective validation.
ER -