Hip joint centre localisation with an unscented Kalman filter

Area of Science:

• Biomechanics
• Orthopaedic Surgery
• Robotics

Background:

• Accurate hip joint centre (HJC) estimation is fundamental for gait analysis and computer-assisted orthopaedic procedures.
• Current functional methods often rely solely on femur kinematics, potentially leading to errors due to unobserved pelvic displacements.

Purpose of the Study:

• To design and test an unscented Kalman filter for improved HJC estimation.
• To incorporate pelvic marker data to correct for spatial displacements during the pivoting maneuver.

Main Methods:

• Development of an unscented Kalman filter integrating femur pose and single pelvic marker 3D coordinates.
• Validation using a hip joint mechanical simulator with realistic tissue properties.
• Comparison of algorithm performance against established literature standards.

Main Results:

• The proposed unscented Kalman filter demonstrated enhanced performance in HJC estimation.
• Superior accuracy was observed in cases with pelvic translation exceeding 8 mm.
• The algorithm meets the clinical requirements for precision in orthopaedic applications.

Conclusions:

• The novel unscented Kalman filter effectively estimates the hip joint centre by integrating femur and pelvic motion.
• This method offers improved accuracy, particularly in the presence of pelvic translation, addressing limitations of previous techniques.
• The validated approach holds significant potential for enhancing precision in gait analysis and computer-assisted orthopaedic surgery.