3D Patellar Shape is Associated with Patellar Dislocation: an Automated Coordinate Algorithm and Statistical Shape...

Area of Science:

• Orthopedics
• Biomedical Engineering
• Medical Imaging

Background:

• Patellar dislocation (PD) is a common orthopedic issue.
• Understanding the 3D morphology of the patella is crucial for diagnosing and treating PD.
• Existing methods for patellar analysis lack standardization and quantitative precision.

Purpose of the Study:

• To develop an automated, landmark-based coordinate system for the patella.
• To create a patellar statistical shape model (SSM) for analyzing 3D morphology.
• To identify and quantify 3D morphological variations associated with patellar dislocation (PD).

Main Methods:

• Reconstructed patellar surface models from CT/MRI scans of 54 participants (33 PD, 21 controls).
• Established and validated an automated patellar coordinate system.
• Utilized principal component analysis (PCA) to build an SSM and extract major 3D shape modes.
• Assessed demographic/morphometric risk factors and evaluated a logistic regression classifier.

Main Results:

• The automated coordinate system demonstrated high repeatability.
• No significant differences in patellar linear dimensions or centroid size were found between PD and control groups.
• Two principal components (PC4 and PC7) representing thickness/facet and facet-edge morphology, respectively, significantly differentiated PD from controls.
• A cross-validated classifier achieved good in-cohort discrimination (mean AUC ≈ 0.91).

Conclusions:

• Patellar dislocation is associated with distinct 3D articular surface shape patterns, including a prominent medial facet and flattened posterolateral facet, independent of overall patellar dimensions.
• The automated coordinate system and SSM provide a reproducible method for quantitative patellar phenotyping.
• Identified shape modes offer insights into PD pathomechanics and a basis for future risk modeling.