Explainable machine learning for osteoporosis detection in patients with osteopenia: model development and validation using routine clinical data from an Asian cohort - Report - MDSpire
Advertisement
Development and Validation of an Explainable Machine Learning Model for Osteoporosis Identification in Osteopenic Patients Using Routine Clinical Data from an Asian Population
Clinical Report: Explainable Machine Learning Model for Osteoporosis Identification
Overview
This study developed a machine learning model to distinguish osteoporosis from osteopenia using routine clinical data in an Asian population. The model demonstrated a mean AUC of 0.738.
Background
Osteoporosis is a significant public health issue, particularly in Asia, where access to diagnostic tools like DXA is limited. Osteopenia, a precursor to osteoporosis, affects a substantial portion of the population, yet current screening methods are inadequate. This study addresses the need for accessible screening tools that utilize routine clinical data to identify individuals at risk.
Data Highlights
Model
Mean AUC
Independent Validation AUC
Key Predictors
Linear Discriminant Analysis
0.738 (95% CI: 0.736–0.741)
0.710 (95% CI: 0.686–0.734)
Waist-to-height ratio, body weight, serum uric acid, age, alkaline phosphatase
Key Findings
The Linear Discriminant Analysis model outperformed other algorithms in distinguishing OP from osteopenia.
The model achieved a mean cross-validated AUC of 0.738 and maintained an AUC of 0.710 on independent validation.
Key predictors for the model included waist-to-height ratio, body weight, serum uric acid, age, and alkaline phosphatase.
Decision curve analysis indicated a positive net benefit across various risk thresholds.
Clinical Implications
The developed machine learning model provides a practical approach for identifying individuals at risk of osteoporosis using readily available clinical data.
Conclusion
The study presents a viable machine learning model for osteoporosis screening. Further validation in diverse populations is needed.
Teriparatide followed by zoledronic acid increased bone mineral density but did not reduce fracture risk compared with standard care in adults with osteogenesis imperfecta.
