AI to estimate BMD from Opportunistic screening of X-rays
Research type
Research Study
Full title
A prospective, cross-sectional study to assess the performance of the opportunistic osteoporosis screening tool, OsteoSight, in quantifying bone mineral density (BMD) from anteroposterior (AP) hip and pelvic X-rays in over 50s.
IRAS ID
342387
Contact name
Cat Kelly
Contact email
Sponsor organisation
Naitive Technologies
Clinicaltrials.gov Identifier
10084038, Innovate UK Advancing Precision Medicine Award
Duration of Study in the UK
0 years, 8 months, 29 days
Research summary
This pilot study aims to assess the performance of OsteoSight, a new tool for identifying those at risk of osteoporosis, from routine X-rays. Osteoporosis is currently diagnosed using DXA scans, which provide a measure of Bone Mineral Density (BMD). However, with competing health concerns against a background of chronic under resourcing, many primary care doctors do not routinely order DXA scans, leading to missed opportunities to prevent fragility fractures. OsteoSight offers a potential solution to case-finding by leveraging the thousands of X-ray images taken every day for many reasons. OsteoSight uses machine learning models to automatically provide a BMD estimate from X-ray images. A positive result can then alert the patient's care team to further investigate according to standard of care. To be effective as a screening tool, OsteoSight should perform well in its target population.
The study involves 100 patients, aged 50 years and over, who are undergoing a hip or pelvic X-ray as part of their standard of care at the Royal National Orthopaedic Hospital (RNOH). Each participant will be invited to attend a single visit, where they will have a DXA scan and answer some questions about fracture risk as part of a bone health and awareness assessment. The images obtained from the X-ray scans will be analyzed using OsteoSight to provide an estimate of BMD, with DXA serving as the reference standard against which OsteoSight's performance will be evaluated.
Ultimately, the findings of this study could have significant implications for clinical practice, potentially leading to more widespread and accessible osteoporosis screening programs. This, in turn, could help identify individuals at risk of fractures earlier, allowing for timely interventions to reduce the burden of osteoporosis-related complications.
This study is sponsored by Naitive Technologies, the manufacturers of OsteoSight, and is funded by a Precision Medicine Award from Innovate UK.
REC name
London - Stanmore Research Ethics Committee
REC reference
25/LO/0163
Date of REC Opinion
10 Apr 2025
REC opinion
Further Information Favourable Opinion