Societal Impacts of Opportunistic Detection of Osteoporotic Vertebral Fragility Fractures using Artificial Intelligence: A Literature Review 

Author: Georgina Lindsay

Purpose

Vertebral fragility fractures (VFFs), a hallmark of osteoporosis, are a major global health concern associated with significant morbidity and mortality [1]. Early identification of VFFs enables timely osteoporosis assessment, treatment initiation and reduction in future fracture risk [2, 3]. Despite this, VFFs are chronically underreported with up to 70% of incidental VFFs on imaging missed [4]. Artificial intelligence (AI) offers a promising opportunity to address this gap by enabling systematic, opportunistic detection across existing imaging [2].

Method

A literature review was conducted to explore the societal impacts of AI-based opportunistic VFF detection.

Results

A range of AI-based technologies for automated VFF detection have been developed. Early pilots indicate strong performance with sensitivity and specificity that varies among different models, but remains high across most studies [5, 6, 7]. This suggests AI models may improve accuracy and efficiency of diagnosing VFFs in clinical practice. It may also reduce reporting variation, particularly in regions without subspecialist radiologists, helping to narrow geographical inequalities in diagnostic quality and reach patients who may otherwise remain undiagnosed [2].

However, there is significant heterogeneity and study biases within different AI models. Models may be influenced by different variables such as imaging protocols and fracture severity, raising concerns about false positives and potential over investigations [5, 8]. Furthermore, there is limited transparency about characteristics of patient populations in training datasets in the clinical evidence, with concerns about underrepresentation of demographic groups including younger people, ethnic minorities, people with comorbidities, atypical morphology, or previous treatments [2]. Without rigorous external validation, these limitations may inadvertently reinforce existing inequalities [5, 9].

The impact on treatment and outcome of osteoporosis patients has not been demonstrated yet due to lack of prospective studies. However, the value of AI-enabled VFF detection will depend on robust referral pathways and downstream fracture management services [2, 3, 10]. Increased detection rates may overwhelm referral pathways, which already face capacity and geographical variation [11]. Without proportional investment in follow-up services, increased detection may inadvertently exacerbate inequalities [2, 3]. Conversely, AI-derived data could illuminate the true unmet need and support equitable planning and expansion of fracture liaison services in underserved regions [2].

Conclusion

Opportunistic detection of VFFs using AI is potentially transformative for patients. The technology could significantly reduce the societal burden of osteoporosis by enabling proactive and effective case finding and instigate appropriate referral and treatment to prevent further fractures and associated morbidity and mortality. However, it requires diverse training data sets, robust workflow integration, and significant investment in downstream care pathways to ensure safe, effective, and equitable deployment.

References

[1] Clinical Guidance for the Effective Identification of Vertebral Fractures, National Osteoporosis Society, 2022

[2] Artificial intelligence (AI) technologies to aid opportunistic detection of vertebral fragility fractures: early value assessment, NICE Guidance, 2025

[3] Adding value in radiology—improved radiological diagnosis of osteoporotic vertebral fragility fractures following National UK Audit and Interventions, Howlett et al., 2025

[4] Radiology reporting of incidental osteoporotic vertebral fragility fractures present on CT studies: results of UK national re-audit, Howlett et al., 2023

[5] Artificial Intelligence in Risk Prediction and Diagnosis of Vertebral Fractures, Namireddy et al., 2024

[6] Clinical Validation of Commercial AI Software for the Detection of Incidental Vertebral Compression Fractures in CT Scans of the Chest and Abdomen, Mathew et al., 2025

[7] Using Artificial Intelligence to Diagnose Osteoporotic Vertebral Fractures on Plain Radiographs, Shen et al., 2023

[8] Breaking the silence: AI’s contribution to detecting vertebral fractures in opportunistic CT scans in the elderly—a validation study, Spangeus et al., 2025

[9] AI and the quest for diversity and inclusion: a systematic literature review, Sham et al., 2025

[10] Improved radiological diagnosis of osteoporotic vertebral fragility fractures following UK-wide interventions and re-audit-can this be maintained and translated into clinical practice? Adams et al., 2025

[11] Integration of a vertebral fracture identification service into a fracture liaison service: a quality improvement project, Ong et al., 2021