Artificial intelligence in osteoporosis assessment using CT imaging: a scoping review

Area of Science:

• Radiology and Medical Imaging
• Artificial Intelligence in Healthcare
• Osteoporosis Research

Background:

• Osteoporosis assessment traditionally relies on dual-energy X-ray absorptiometry (DXA) and quantitative CT (QCT).
• Computed Tomography (CT) imaging offers a potential alternative for opportunistic osteoporosis assessment.
• The integration of Artificial Intelligence (AI) presents new avenues for analyzing CT data for bone health.

Purpose of the Study:

• To systematically review and map current research on AI applications in CT-based osteoporosis assessment.
• To focus on the methodological approaches, anatomical regions, and algorithmic performance of AI in this field.
• To identify trends and gaps in the existing literature.

Main Methods:

• A comprehensive literature search was conducted across PubMed, EMBASE, and Web of Science databases (1995-2025).
• Studies utilizing AI, machine learning, or deep learning on CT images for osteoporosis classification, bone mineral density (BMD) estimation, or fracture-risk prediction were included.
• Data extraction focused on study characteristics, imaging sources, analytical workflows, and validation methods.

Main Results:

• 51 studies were included, predominantly retrospective and single-center, with many originating from China.
• AI models demonstrated high performance for osteoporosis diagnosis (AUC 0.80-0.997) and opportunistic screening (AUC 0.781-0.99).
• Fracture-risk prediction models showed more variable accuracy (AUC 0.702-0.92), and significant methodological heterogeneity was observed across studies.

Conclusions:

• AI-enhanced CT shows diagnostic and screening performance comparable to established methods like DXA and QCT.
• Methodological diversity and inconsistent validation strategies currently limit the generalizability and clinical translation of AI tools.
• Standardizing AI workflows and integrating multimodal data could enhance fracture-risk prediction and facilitate widespread clinical adoption for osteoporosis management.