2025 ISAKOS Biennial Congress Paper
Validation of Robotically-Assisted Patella Tracking in Total Knee Arthroplasty: An in Vitro Biomechanics Study
Jobe Shatrov, MD, St Leonards, NSW AUSTRALIA
Elizabeth Clarke, PhD, sydney, nsw AUSTRALIA
Koki Abe, PhD, yokohama, Japan JAPAN
Mounir Boudali, PhD, sydney, nsw AUSTRALIA
Bill L Walter, Prof, PhD, MBBS, St Leonard, NSW AUSTRALIA
David A. Parker, MBBS, BMedSc, FRACS, Sydney, NSW AUSTRALIA
Kolling Institute, Northern Clinical School, Faculty of Medicine and Health, University of Sydney, Sydney, st leonards, nsw, AUSTRALIA
FDA Status Not Applicable
Summary
This study demonstrates that the described robotically-assisted technique allows for precise visualization of the patella flexion arc, which correlates with changes in patellar motion, joint reaction forces, and quadriceps tendon strain.
Abstract
Objective
The intra-operative assessment of patella tracking during total knee arthroplasty (TKA) has evolved little since being described in 1979. Robotic-tools have been developed to improve planning, accuracy and outcomes in TKA. The purpose of this study was to test the validity of a recently described technique for assessing the patellofemoral (PFJ) in TKA using an imageless robotic platform.
Methods
This biomechanical study utilized cadaveric knees, maintaining native muscle attachments from the pelvis to the knee. TKA was performed using the CORI robotic system (Smith & Nephew) using Legion II components and a posterior-stabilized tibial insert via medial parapatellar approach. Femoral and tibial resection depths included cartilage thickness, and patella resurfacing was conducted using a custom dome-shaped prosthesis to measure pressures across three patellar regions. Kinematics were tracked with the OptiTrack motion capture system.
The SimVITRO (Cleveland clinic, OH) musculoskeletal simulator was used to test loading conditions in the knee joint through coordinated control of an external loading device (rotopod). Controlled loading conditions were simulated utilising a hybrid of knee flexion position control and real-time proportional-integral-derivative (PID) force feedback control to simulate passive knee flexion and descending stairs. The forces trajectories were adapted from previously published in vivo testing data. The simulated conditions tested the knee through a range of motion of 0°- 100° flexion.
Cadaveric knees were registered using the CORI (Smith and Nephew) robotic knee system and the patella resurfaced using a custom designed dome shaped prosthesis to create a variety of thicknesses compared to the native state (+4mm, neutral, oblique and -3mm). Kinematics were recorded using the Optitrak motion capture system.
Combined motion capture, soft tissue strain and patella joint reaction forces were recorded during testing using a previously validated orthopaedic biomechanical testing machine (SIMVitro robot). The knee was tested in the following states; passive knee flexion, sit-to stand and descending stairs through a range of motion of 0°- 100° flexion. The patella tracking was recorded using the CORI and SIMVitro systems and the data analysed.
Results
The robotic motion capture system recorded changes in the patella flexion arc that corresponded directly to the changes in the thickness of the patella. Motion patterns recorded by the CORI system were correlated and verified with the motion capture system. The native patella had a flexion radius of 54.8 mm. When the patella was over-stuffed by 4mm compared to the native state the radius increased to 59.7 mm. The patella tracked further laterally with subsequent increases in thickness. Patella joint reaction forces demonstrated a direct relationship with changes to the patella flexion arc captured by the CORI system. Soft tissue strain in the quadriceps tendon demonstrated a direct relationship with changes to the patella flexion axis. Under-stuffing the patella by 3mm led to loss of strain the in the quadriceps tendon through mid-flexion range (~20° - 60° flexion). An obliquely resurfaced patella exhibited significant deviation in both the patella flexion arc but also the patella tracking from the native state.
Conclusion
This study demonstrates that patella tracking can be accurately assessed using a commercially available robotic platform. The novel technique allows for precise visualization of the patella flexion arc, which correlates with changes in patellar motion, joint reaction forces, and quadriceps tendon strain. These findings have important implications for enhancing intra-operative patella tracking assessment during TKA.