Prediction of Rotator Cuff Muscle Fibre Orientations Using a Population-Averaged Atlas Generated With Anatomical and...

Area of Science:

• Biomechanics
• Medical Imaging
• Anatomy

Background:

• Accurate in vivo muscle architecture measurements are challenging.
• Understanding rotator cuff muscle function requires detailed anatomical data.
• Existing methods for muscle fiber orientation analysis have limitations.

Purpose of the Study:

• To develop population-averaged atlases of human rotator cuff muscle shape and fiber orientations.
• To utilize these atlases for predicting muscle fiber orientations from anatomical MRI alone.
• To validate the accuracy of atlas-based predictions using diffusion-weighted imaging (DWI).

Main Methods:

• An image registration framework was used to coregister anatomical MRI and DWI data from 20 subjects into sex-specific common spaces.
• Population-averaged atlases were created for male and female rotator cuff muscles.
• Leave-one-out cross-validation was performed to assess prediction accuracy using Dice coefficients, ACCs, and angular differences.

Main Results:

• Individual MRI and DWI data were accurately registered into sex-specific atlas spaces.
• The atlases enabled accurate prediction of muscle fiber orientations from anatomical MRI alone.
• Predicted orientations closely matched DWI-derived data, showing improved smoothness and coverage.

Conclusions:

• Population-averaged atlases significantly enhance muscle architecture reconstructions.
• Anatomical MRI alone is sufficient for accurate prediction of rotator cuff muscle fiber orientations in healthy young individuals.
• This approach offers a non-invasive method for detailed muscle analysis.