Impact of Marker-set and Joint Constraint on kinematic outcome in Patients with knee Osteoarthritis

Gait analysis is crucial for the assessment of lower limb motor disorders, yet the optimal model for accurately capturing complex knee joint kinematics remains a research subject and is particularly underexplored for patients with osteoarthritis. Optoelectronic motion capture systems, currently considered the gold standard, can quantitatively assess 3D knee kinematics, including knee flexion/extension, abduction/adduction, and internal/external rotation across the sagittal, frontal, and transverse planes. This study aims to assess the agreement of two prevalent lower limb gait models, the Istituto Ortopedico Rizzoli (IOR) and the Cluster-Based Anatomical System Technique (CAST), employing 6DoF and inverse kinematics. The inter-model agreement was assessed through the intraclass correlation coefficient (ICC) and mean absolute differences (MD) to analyze knee joint dynamics. High consistency was observed in sagittal plane measurements (ICC $\geq \mathbf{0. 7 5}$) across all models and gait phases. However, substantial variability was noted in the frontal and transversal planes, with the greatest discrepancies in knee intra- and extra-rotation. Differences between models were most pronounced in the frontal plane between the 6DoF configurations of the CAST and IOR models, with absolute mean differences up to 12.7° during maximum swing. Minor discrepancies were observed between Inverse kinematic (IK) configurations of the CAST and IOR models. While both IOR and CAST models demonstrated robust capabilities for assessing sagittal plane dynamics, challenges persist in accurately capturing complex joint motions outside this plane, which is essential for comprehensive gait analysis in osteoarthritis. The study highlights the importance of precise selection and customization of gait models based on specific clinical or research requirements to enhance the reliability of kinematic data.