Anatomical frame identification and reconstruction for repeatable lower limb joint kinematics estimates

Area of Science:

• Biomechanics
• Kinesiology
• Medical Imaging

Background:

• Accurate joint mechanics description relies on precise anatomical axis reconstruction.
• Traditional anatomical calibration methods often involve manual landmark palpation, leading to indeterminacy.
• Repeatability of anatomical calibration is critical for reliable kinematic and kinetic analyses.

Purpose of the Study:

• To present an automated anatomical calibration approach for estimating subject-specific bone morphology.
• To identify anatomical frames automatically, reducing reliance on manual palpation.
• To assess the repeatability of joint angle measurements using the proposed method.

Main Methods:

• Photogrammetric digitization of superficial bone points.
• Matching digitized data with a deformable template bone model.
• Incorporating functional joint center data for enhanced accuracy.

Main Results:

• The method demonstrated limited dispersion (<3 deg) for most joint angles in upright posture.
• Hip and knee internal-external rotation showed higher dispersion (6-9 deg) in upright posture.
• During walking, inter-operator and inter-trial errors were comparable for pelvic, hip, and knee flexion-extension angles (0.1-0.9 deg).
• Knee internal-external rotation and abduction-adduction exhibited slightly higher inter-operator errors (0.9-2.9 deg).

Conclusions:

• The proposed anatomical calibration method offers improved repeatability for joint angle measurements.
• Automated identification of anatomical frames enhances objectivity and reduces operator dependency.
• This technique provides a more reliable foundation for quantitative joint mechanics analysis.