Prospective validation of a rapid CT-based bone mineral density screening method using colored spinal images

Area of Science:

• Radiology
• Medical Imaging
• Bone Health

Background:

• Accurate assessment of bone mineral density (BMD) is crucial for diagnosing osteoporosis.
• Traditional methods for BMD assessment can be time-consuming or require specialized equipment.

Purpose of the Study:

• To prospectively validate a novel method for differentiating normal from abnormal spinal bone mineral density (BMD).
• To assess the accuracy and speed of using colored abdominal CT images for BMD evaluation.

Main Methods:

• 196 asymptomatic women aged 50+ underwent routine nonenhanced abdominal CT scans.
• CT images were processed to generate sagittal colored views, with green indicating normal BMD and red indicating abnormal BMD.
• Radiologists evaluated BMD using visual assessment of grayscale and colored images, and quantitative measurements, compared against a phantom-based reference standard.

Main Results:

• Colored CT image assessment achieved 92% accuracy in 2.1 seconds, significantly outperforming grayscale visual assessment (66% accuracy in 6.0 seconds) and quantitative measurements (88% accuracy in 15.2 seconds).
• Inter-observer agreement was excellent (ICC: 0.90) for the colored image method, compared to poor (ICC: 0.31) for grayscale and good (ICC: 0.73) for quantitative measurements.
• Mean attenuation on CT strongly correlated with BMD (r=0.96/0.91).

Conclusions:

• Colored abdominal CT imaging provides a highly accurate and rapid method for detecting abnormal BMD.
• This technique demonstrates excellent inter-observer agreement, improving diagnostic reliability.
• The method offers a promising, efficient approach for osteoporosis screening and diagnosis.