Automated skeletal tissue quantification in the lower leg using peripheral quantitative computed tomography.

Area of Science:

• Biomedical Engineering
• Radiology
• Orthopedics

Background:

• Accurate quantification of bone and soft tissue properties is vital for understanding skeletal health and mechanical load.
• Assessing muscle volume, density, and fat infiltration provides insights into the muscle-bone relationship.

Purpose of the Study:

• To introduce a novel methodology for automated quantification of hard and soft tissues at multiple tibia sites.
• To develop a user-friendly tool for clinical researchers to analyze bone and muscle characteristics.

Main Methods:

• Utilized implicit active contour models and clustering for automated segmentation of bone, muscle, and fat.
• Calculated densitometric, area, and shape characteristics for each tissue type.
• Implemented the approach as a multi-platform tool named TIDAQ (tissue identification and quantification).

Main Results:

• Validated the method against reference measurements, showing high agreement (Deming regression slopes ~1.1) for areas and densities.
• Achieved high coefficients of variation (R² > 0.88) for cross-sectional areas and mineral densities.
• Demonstrated excellent tissue segmentation accuracy with an average Dice coefficient of 0.91.

Conclusions:

• The developed methodology enables high-throughput, accurate, and reproducible automatic quantification of lower leg bone and muscle characteristics.
• This information is critical for evaluating adverse outcome risks and assessing interventions for bone and muscle strength.