Quantifying Subresolution 3D Morphology of Bone with Clinical Computed Tomography

Area of Science:

• Biomedical Engineering
• Radiology
• Orthopedics

Background:

• Trabecular bone morphometrics are crucial for understanding osteoarthritis (OA) but are difficult to quantify with clinical imaging.
• Sub-resolution details in bone structure are linked to OA severity.
• Cone beam computed tomography (CBCT) offers potential for in vivo assessment of bone microarchitecture.

Purpose of the Study:

• To quantify sub-resolution trabecular bone morphometrics from clinical resolution CBCT images.
• To evaluate the association between texture/histogram parameters and bone morphometrics.
• To develop a statistical model for estimating bone morphometrics using CBCT data.

Main Methods:

• Human tibial and femoral bone samples were imaged using CBCT.
• Grey-level co-occurrence matrix (GLCM) texture and histogram parameters were calculated.
• Parameters were correlated with micro-CT quantified morphometrics and OARSI histopathological grading for OA severity.

Main Results:

• Individual histogram and GLCM parameters strongly correlated with bone morphometrics (e.g., histogram mean and bone volume fraction, r=0.907).
• A combined statistical model of GLCM and histogram parameters improved the estimation of bone volume fraction from CBCT data.
• Histopathology showed moderate associations with bone morphometrics.

Conclusions:

• CBCT-imaged trabecular bone texture and histogram parameters are associated with sub-resolution morphometrics.
• Sub-resolution bone morphometrics can be estimated from clinical CBCT images.
• Combining histogram and GLCM parameters enhances the accuracy of these estimations.