Correspondence model-based approach for evaluating static and dynamic joint distance measurements.

Area of Science:

• Biomedical Engineering
• Computational Anatomy
• Medical Imaging

Background:

• Traditional 3D joint space measurements use surface divisions, limiting shape impact analysis.
• Morphometric analyses and shape variations are crucial for understanding disease progression and deformity.
• Correspondence models offer co-registered landmarks for analyzing joint structure variability.

Purpose of the Study:

• To propose and validate a novel method for group-wise statistical analysis of joint space measurements using correspondence models.
• To verify and validate the measurement accuracy and reliability of a developed open-source toolbox.
• To assess joint space variability in static or quasi-static positions during movement.

Main Methods:

• Utilized correspondence model-based shape analyses with co-registered landmarks.
• Performed group-wise statistical analyses on joint space measurements.
• Tested an open-source toolbox with surface meshes of varying edge lengths (0.5–2 mm) and joint space distances (1–4 mm).

Main Results:

• Achieved <1% error for 0.5- and 1-mm mesh edge lengths with 4 mm joint space.
• Demonstrated best sensitivity with a 0.5 mm mesh edge length.
• Confirmed repeatable and reliable measurements with low coefficient of variation and high intraclass correlation coefficient.

Conclusions:

• Correspondence model-based approaches provide robust and accurate joint measurements, overcoming limitations of traditional methods.
• The developed open-source toolbox offers a practical tool for assessing joint-level disease and deformity by incorporating shape variability.
• Future work will explore clinical applications, including highly deformed joint structures.