Univ Lyon, Univ Gustave Eiffel, LBMC UMR_T 9406, F-69622 Lyon, France.
Willy Taillard Laboratory of Kinesiology, University Geneva Hospitals and Geneva University, Geneva, Switzerland.
J Biomech. 2020 Dec 2;113:110117. doi: 10.1016/j.jbiomech.2020.110117. Epub 2020 Nov 9.
Accurate estimation of the tibiofemoral contact forces relies on exact kinematics and joint geometry. Subject-specific kinematic constraints representing contact point trajectories derived from fluoroscopic measurements during lunge are introduced in a musculoskeletal model of the lower limb and compared to generic kinematic constraints. The medial, lateral, and total contact forces during gait and squat are validated using the data of four patients with an instrumented prosthesis. The accuracy of the estimated contact forces (both with subject-specific and generic kinematic constraints) remains close to the level reported in the literature. The mean root mean square errors range from 0.32 to 0.52 body weights for gait and from 0.27 to 0.72 body weights for squat. The impact of the subject-specific contact point trajectories is not found substantial or consistent between patients and tasks. Indeed, the kinematics of the total knee prostheses remains close to the kinematics of a hinge joint and the contact point locations remain generally centred at 20 mm from the tibia centreline (close to the constant value defined in the generic constraints). The contact point trajectories are also suspected to differ between tasks (lunge vs. gait and squat). While the contact point trajectories have been reported to be sensitive model parameters, no clear improvement of the contact force accuracy is demonstrated on patients with instrumented prosthesis. The introduction (as kinematic constraints) of fluoroscopy-based contact point trajectories may be considered in cases where these trajectories are significantly altered, as reported for osteoarthritis patients.
准确估计胫股接触力依赖于精确的运动学和关节几何形状。本研究将代表来自伸展位荧光透视测量的接触点轨迹的特定于个体的运动学约束引入到下肢肌肉骨骼模型中，并与通用运动学约束进行了比较。使用带有仪器化假体的 4 名患者的数据验证了步态和深蹲期间的内侧、外侧和总接触力。估计的接触力（使用特定于个体和通用运动学约束）的准确性仍然接近文献报道的水平。平均均方根误差范围为 0.32 到 0.52 体重，用于步态，0.27 到 0.72 体重，用于深蹲。特定于个体的接触点轨迹的影响在患者和任务之间并不明显或一致。实际上，全膝关节假体的运动学仍然接近铰链关节的运动学，并且接触点位置通常保持在距胫骨中心线 20mm（接近通用约束中定义的常数）处居中。接触点轨迹也被怀疑在任务之间存在差异（伸展与步态和深蹲）。虽然接触点轨迹已被报道为敏感的模型参数，但在带有仪器化假体的患者中，接触力准确性没有明显提高。在接触点轨迹明显改变的情况下（如骨关节炎患者），可以考虑将基于荧光透视的接触点轨迹（作为运动学约束）引入。
