Dynamic In Vivo Experimental And Computational Framework For Tka Evaluation Of Lunge And Seiza Kneeling

Download or read book Dynamic, In-Vivo Experimental And Computational Framework For TKA Evaluation Of Lunge And Seiza Kneeling written by and published by . This book was released on 2017 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: INTRODUCTION: Achieving deep knee flexion is essential for quality of life in most Asian cultures, particularly the Japanese population. Seiza is a common kneeling task performed in Japan that achieves deep flexion (>120u2070) with resting thigh-calf contact. To accommodate Seiza and other similar tasks, such as gardening and squatting, implant designers are interested in evaluating healthy and implanted knee mechanics during deep knee flexion activities. Total knee replacement (TKR) designs are designed to allow for high flexion range of motion and function in harmony with surrounding soft tissue. While previous studies have measured joint kinematics during Seiza using fluoroscopic imaging [1,2], joint contact and soft tissue forces remain unknown. Our prior work combined experimental and computational approaches with high-speed stereo radiography, musculoskeletal modeling, and finite element (FE) analysis to develop a subject-specific, load-controlled simulation of the lunge activity in a healthy subject [3]. The goal of the current study was to expand on our prior work to develop a predictive simulation of Seiza for the same subject, and to further validate the model for evaluation of TKR-implanted mechanics. The load-controlled simulation was used to compare healthy and implanted knee mechanics during Seiza kneeling.METHODS: Simultaneous high-speed stereo radiography (HSSR) images, marker-based motion capture, and ground reaction forces were collected for a healthy, older adult male (age=52years, height=172cm, weight=126lbs, BMI=19.3) performing Seiza kneeling [4] (Figure 1a). The relative position of femur, tibia, and patella bones were tracked from the HSSR images using Autoscoper (Brown University, Providence, RI) by registering CT-reconstructed bones to the 2D HSSR images. Tibiofemoral (TF) and patellofemoral (PF) joint kinematics were described relative to full extension [5]. A previously, validated FE model of the healthy subject, including subject-specific calibrated soft tissue properties, was virtually implanted with TKR posterior-stabilized components (Scorpio NRG, Stryker) using mechanical-axis alignment (0u2070 varus and tibial slope) and guidance from an orthopaedic surgeon [3] (Figure 1b). The FE model included bone, cartilage/TKR implants, TF (1D, non-linear springs) and PF (2D membranes) ligaments, meniscus (3D hexahedral elements and 1D linear horn and periphery attachments), and quadriceps and hamstrings muscles (1D connectors). Model loading and boundary conditions included a combination of hip, ankle, and muscle (e.g. PID-controlled quadriceps force) loads derived from HSSR kinematics and musculoskeletal models. Model loads were derived from simulation of the natural condition and were then applied to the implanted model. While the natural model was previously validated using direct comparison to experimental subject-specific knee kinematics in the lunge activity, the virtually implanted model was validated using comparison to fluoroscopy-based TF kinematics from the literature [1]. Model loading conditions were developed for load-controlled simulation of the Seiza kneeling activity. While TF flexion was kinematically prescribed based on HSSR kinematics, TF varus-valgus (V-V), internal-external (I-E), anterior-posterior (A-P) and superior-inferior (S-I) motion were force-driven using PID-controlled joint loads, calibrated to match the natural experimental in-vivo motion during Seiza. As in the lunge activity, predicted healthy loads were applied to the implanted condition. Load-controlled simulation of Seiza was used to predict implanted TF kinematics, joint contact and ligament forces. RESULTS: The subject achieved peak flexion of 132u00b0 during lunge, and peak flexion and I-E rotation of 151u00b0 and 30u00b0, respectively, during Seiza. In lunge simulation, the implanted model was consistent with fluoroscopy kinematics from the literature in both trend and magnitude (Figure 2a). The medial compartment distracted during Seiza for both natural and implanted conditions (peak 7u00b0 valgus). In simulation of Seiza, the calibrated natural model replicated HSSR-based kinematics (RMS diff.