Opportunistic osteoporosis screening using chest CT with artificial intelligence

Area of Science:

• Radiology and Imaging
• Bone Health
• Artificial Intelligence in Medicine

Background:

• Osteoporosis is a prevalent condition with a low detection rate, leading to fractures.
• Early detection is crucial for preventing severe consequences like osteoporotic fractures.

Purpose of the Study:

• To evaluate the utility of thoracic and first lumbar vertebral attenuation values, measured by AI on non-enhanced chest CT, for osteoporosis screening.

Main Methods:

• AI measured attenuation values (HU) of thoracic and lumbar vertebrae from chest CT scans.
• Patients were categorized into normal, osteopenia, and osteoporosis groups based on DXA results.
• Correlation between CT attenuation values and BMD, predictive ability, and diagnostic efficacy were analyzed.

Main Results:

• Vertebral CT values decreased with age, particularly in postmenopausal women.
• Increased CT values showed significant inverse correlation with osteopenia and osteoporosis risk.
• Combined analysis of all thoracic vertebrae yielded higher diagnostic efficacy (AUC 0.831-0.972) than single vertebrae.

Conclusions:

• Routine chest CT with AI enables opportunistic screening for osteopenia and osteoporosis.
• This method efficiently identifies high-risk individuals without additional radiation exposure.
• Improved osteoporosis detection can reduce the incidence of osteoporotic fractures.