Assessment of trabecular structure using high resolution CT images and texture analysis

Area of Science:

• Biomedical Engineering
• Radiology
• Orthopedics

Background:

• Trabecular bone structure is crucial for bone strength.
• Assessing bone strength non-invasively is essential for fracture risk prediction.
• High-resolution CT (HR CT) offers detailed imaging of bone microarchitecture.

Purpose of the Study:

• To investigate human vertebral trabecular structure using HR CT and texture analysis.
• To compare texture analysis and bone mineral density (BMD) in predicting bone strength.
• To determine if combining texture analysis with BMD enhances bone strength prediction.

Main Methods:

• HR CT images (1 mm slice thickness) were acquired from 28 human vertebral bone cubes.
• Four texture analysis techniques were applied to the HR CT images.
• Quantitative CT (QCT) was performed, and elastic modulus (EM) was measured biomechanically.

Main Results:

• A significant correlation was found between elastic modulus (EM) and BMD (R2 = 0.78).
• Texture measures from HR CT images also showed significant correlations with EM.
• A multivariate regression model combining texture measures and BMD achieved a higher prediction accuracy for EM (R2 = 0.87).

Conclusions:

• Texture parameters derived from HR CT images are significantly correlated with the elastic modulus of human vertebral bone.
• Integrating texture analysis with BMD measurements provides a more robust prediction of bone strength compared to BMD alone.
• This combined approach holds promise for improved non-invasive assessment of bone strength.