Introducing Anisotropic Minkowski Functionals and Quantitative Anisotropy Measures for Local Structure Analysis in...

Area of Science:

• Biomedical Imaging
• Medical Image Analysis
• Quantitative Structure Analysis

Background:

• Minkowski Functionals are established for characterizing local structure in medical imaging.
• Existing methods lack the ability to capture the inherent anisotropy of gray-level structures.

Purpose of the Study:

• Introduce Anisotropic Minkowski Functionals (AMFs) to quantify structural anisotropy.
• Apply AMFs to trabecular bone micro-architecture in the proximal femur using multi-detector CT.
• Evaluate AMFs' potential in predicting bone strength.

Main Methods:

• Computed local AMFs for regions of interest (ROIs) in the femoral head, neck, and trochanter.
• Utilized fractional anisotropy to quantify local anisotropy of trabecular structures.
• Employed multi-regression models with AMFs to predict bone failure load.

Main Results:

• Demonstrated significantly higher anisotropic trabecular structures in the femoral head and neck compared to the trochanter (p < 10).
• AMFs outperformed bone mineral density (BMD) in predicting failure load when used with multi-regression models.
• AMFs effectively captured directional attributes of local trabecular bone structure.

Conclusions:

• Anisotropic Minkowski Functionals provide valuable insights into directional structural properties.
• AMFs offer a promising approach for enhanced bone strength assessment in biomedical imaging.
• This method has broad applicability across various medical imaging applications.