Quantifying anisotropy of trabecular bone from gray-level images

Area of Science:

• Biomedical Engineering
• Orthopedic Research
• Medical Imaging Analysis

Background:

• Vertebral bone strength is crucial for fracture risk assessment.
• Quantifying bone's structural anisotropy is challenging but important for understanding mechanical properties.
• Existing methods for anisotropy quantification have limitations.

Purpose of the Study:

• To develop and validate a gray-level image-based method for quantifying trabecular bone structural anisotropy.
• To assess the predictive power of anisotropy measures, combined with bone mineral density (BMD), for vertebral mechanical strength (secant modulus).

Main Methods:

• Secant modulus of 30 L(3) vertebral bodies was measured using nondestructive testing.
• Quantitative CT (QCT) was used for BMD measurements.
• Structural anisotropy was quantified using the mean intercept length (MIL) and gray-level structure tensor (GST) methods on binarized and gray-level images, respectively.

Main Results:

• BMD alone explained 28% of the variation in secant modulus.
• Combining BMD with anisotropy measures significantly improved secant modulus prediction.
• The highest correlation (R(2)=0.81) was achieved using BMD and the third principal value of the GST.
• Including minimal cross-sectional area further increased prediction accuracy to 86%.

Conclusions:

• Gray-level image analysis, particularly the GST method, provides valuable insights into bone structural anisotropy.
• Integrating anisotropy measures with BMD offers a more comprehensive approach to predicting vertebral mechanical strength.
• This approach holds potential for improved diagnosis and treatment strategies for skeletal diseases.