Automated Quantitative Bone Analysis in In Vivo X-ray Micro-Computed Tomography

Area of Science:

• Biomedical Engineering
• Osteology Research
• Medical Imaging Analysis

Background:

• Accurate bone morphometry analysis is crucial for preclinical osteological research.
• Automating segmentation and separation of bones in 3D micro-computed tomography (micro-CT) data is challenging due to partial volume effects and thin joints.
• Existing methods struggle with precise bone separation in murine models.

Purpose of the Study:

• To present novel hybrid splitting filters for automated bone separation in 3D micro-CT volumes.
• To introduce a framework for automatically guiding stereological measurements of long bones using their medial axis.
• To enhance the accuracy, consistency, and speed of bone morphometry analysis in preclinical research.

Main Methods:

• Development of hybrid splitting filters using rotationally invariant second-derivative operators to enhance joint contrast.
• Application of marker-controlled watershed segmentation seeded by filter-generated components.
• Characterization of the bone medial axis via morphological thinning and centerline operations.
• Framework for guiding stereological measurements along the bone's longitudinal axis.

Main Results:

• Successfully automated the separation of individual bones, including epiphysis and metaphysis, and removal of growth plates.
• Improved accuracy and consistency in stereological measurements, especially for curved bones like the tibia.
• Demonstrated the efficacy of the combined automated workflow in various in vivo micro-CT studies.

Conclusions:

• The novel hybrid filters and medial axis framework significantly overcome challenges in automated bone separation and morphometric analysis.
• The streamlined software workflow enhances the reliability and efficiency of preclinical osteological research using 3D micro-CT.
• These advancements contribute to more accurate and reproducible bone morphometry measurements.