Mapping AI in osteoporosis screening: a bibliometric analysis of 371 publications

A bibliometric analysis of 371 publications shows AI research in osteoporosis screening and fracture risk prediction surged from 15 papers in 2019 to 86 in 2025, with four knowledge clusters spanning risk assessment, modeling, guidelines, and machine learning.

• 74 co-cited references formed four knowledge-base clusters: early fracture risk assessment, fracture risk model construction, guideline-based management, and machine learning expansion.
• Author keywords (70) revealed four themes: aging-related screening frameworks, ML risk stratification in older fracture populations, imaging/radiomics markers of fragility, and deep learning for detection and triage.
• Top contributors: China (126 publications) and the United States (90).

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.