Integrating subject-specific tibio-femoral contact point trajectories from 3D/2D registration techniques into musculoskeletal model of the lower limb

Ali Zeighami (1, 2) , Rachid Aissaoui (2, 1) , Raphaël Dumas (3)

1 LIO - Laboratoire de recherche en imagerie et orthopédie [Montréal]
2 CHUM - Centre Hospitalier de l'Université de Montréal
3 LBMC UMR T9406 - Laboratoire de Biomécanique et Mécanique des Chocs

Introduction The estimation of the knee joint medial and lateral contact forces depends highly on the location of tibio-femoral Contact Points (CPs) [1]. New imaging techniques revealed high inter-subject variations as well as pathology distinguishable patterns in the trajectories of the CPs in the knee [2]. The objective of this study is to compare the contact forces estimated from a musculoskeletal (MSK) model with standard or personalized CPs. Methods Five pairs of EOSTM low-dose biplanar x-ray images were acquired during a quasi-static squat at 0ᵒ, 15ᵒ, 30ᵒ, 45ᵒ, and 70ᵒ of knee flexion. The tibio-femoral CPs were approximated using the femur and tibia 3D reconstructions and a bone-to-bone proximity algorithm [2]. The 3D gait kinematics and kinetics during gait was used as an input into a MSK model of the lower limb [3] (see figure). The CP trajectories were introduced as kinematic constraints and the corresponding Lagrange multipliers yielded directly the medial and lateral contact forces. The estimated contact forces from the personalized CPs were compared to those obtained using linear CPs which is widely used in most MSK models. Results The difference between medial and lateral contact forces among all the subjects were not statistically significant between the personalized and linear CPs, due to high inter-subject variability. However, in certain subjects, the impact of changing the CP trajectories on the estimated contact forces could reach 1.03 BW and 0.65 BW on the medial and lateral compartments, respectively. Discussion The kinematic constraints based on CP trajectories proposed in this study can be adapted from different joint models and can also integrate measurements from any imaging techniques into the MSK model. This study showed that a medial-lateral shift of the CPs alone (e.g. as in [1]) could not fully predict the changes in the contact forces and that a multifactorial analysis is required, including the anterior-posterior shift as well as the interactions between the musculo-tendon and contact forces. Acknowledgements Fonds de Recherche du Québec en Santé (FRQ-S), Fonds de Recheche du Québec en Nature et Technologie (FRQ-NT), Natural Science and Research Council of Canada (NSERC), and Labex PRIMES (ANR-11-LABX-0063) of Université de Lyon. References [1] Saliba et al. (2017). Sensitivity of medial and lateral knee contact force predictions to frontal plane alignment and contact locations. J Biomech 57: 125-130. [2] Zeighami et al. (2017). Tibio-femoral joint contact in healthy and osteoarthritic knees during quasi-static squat: A bi-planar X-ray analysis. J Biomech 53: 178-184.

Domaines: Biomécanique [physics.med-ph]

Ali Zeighami, Rachid Aissaoui, Raphaël Dumas. Integrating subject-specific tibio-femoral contact point trajectories from 3D/2D registration techniques into musculoskeletal model of the lower limb. 8th World Congress of Biomechanics, Jul 2018, Dublin, Ireland. ⟨hal-02086183⟩