Abstract
Background
The overall goal of total knee arthroplasty (TKA) is to facilitate the restoration of native function following late stage osteoarthritis and for this reason it is important to develop a thorough understanding of the mechanics of a normal healthy knee.
While there are several methods for assessing TKA mechanics, these methods have limitations that make them prohibitive to both replicating physiological systems and evaluating non-implanted knees. These limitations can be circumvented through the development of mathematical models that use anatomical and physiological inputs to computationally simulate joint mechanics. This can be done in an inverse or forward manner to solve for either joint forces or motions respectively. The purpose of this study is to evaluate one such forward model and determine the accuracy of the predicted motions using fluoroscopy.
Methods
In vivo kinematics were determined during flexion from full extension to 120 degrees for ten normal, healthy, subjects using fluoroscopy and a 3D-to-2D registration method. All ten subjects had previously undergone CT scans allowing for the digital reconstruction of native femur and tibia geometries. These geometries were then input into a ridged body forward model based on Kane's system of dynamics. The resulting kinematics determined through fluoroscopy and the mathematical model were compared for all of the ten subjects.
Results
The three kinematic parameters evaluated for this study were the initial positioning and translation of the medial and lateral condylar contact point in addition to the axial position and rotation of the femur with respect to the tibia. The model simulations demonstrated an average of −2.16mm of medial condyle translation, −14.03mm of lateral condyle translation, and 20.09°of axial rotation. Through fluoroscopy, subjects demonstrated an average of −3.63mm of medial condyle translation, −16.02mm of lateral condyle translation, and 15.65°of axial rotation. Comparing these two methods the model predicted on average an additional 1.47mm of medial condyle translation, 1.98mm of lateral condyle translation, and 4.44° less axial rotation compared to the fluoroscopic analysis of the same ten subjects.
Conclusion
In comparing the simulation kinematics to the that of the fluoroscopic assessment, the results are comparably similar demonstrating a forward model can be a viable assessment of knee kinematics in the future. By validating mathematical simulation as a feasible means of mechanical assessment, it becomes possible to evaluate mechanics using inputs to reflect extraordinary and theoretical instances such as trauma patients and congenital deformities unable to be assessed by other methods. The nature of the model also allows for a seamless transition to assess TKA mechanics, creating a more efficient means of evaluating both device design and surgical technique.
