Meta-heuristic optimization algorithms based feature selection for joint moment prediction of sit-to-stand movement using machine learning algorithms.

The sit-to-stand (STS) movement is fundamental in daily activities, involving coordinated motion of the lower extremities and trunk, which leads to the generation of joint moments based on joint angles and limb properties. Traditional methods for determining joint moments often involve sensors or complex mathematical approaches, posing limitations in terms of movement restrictions or expertise requirements. Machine learning (ML) algorithms have emerged as promising tools for joint moment estimation, but the challenge lies in efficiently selecting relevant features from diverse datasets, especially in clinical research settings. This study aims to address this challenge by leveraging metaheuristic optimization algorithms to predict joint moments during STS using minimal input data. Motion analysis data from 20 participants with varied mass and inertia properties are utilized, and joint angles are computed alongside simulations of joint moments. Feature selection is performed using the Manta Ray Foraging Optimization (MRFO), Marine Predators Algorithm (MPA), and Equilibrium Optimizer (EO) algorithms. Subsequently, Decision Tree Regression (DTR), Random Forest Regression (RFR), Extra Tree Regression (ETR), and eXtreme Gradient Boosting Regression (XGBoost Regression) ML algorithms are deployed for joint moment prediction. The results reveal EO-ETR as the most effective algorithm for ankle, knee, and neck joint moment prediction, while MPA-ETR exhibits superior performance for hip joint prediction. This approach demonstrates potential for enhancing accuracy in joint moment estimation with minimal feature input, offering implications for biomechanical research and clinical applications.

Tibio-Femoral Contact Force Distribution of Knee Before and After Total Knee Arthroplasty: Combined Finite Element and Gait Analysis.

OBJECTIVE: To assess the tibio-femoral contact forces before and after total knee arthroplasty (TKA) in patients with knee osteoarthritis (KOA) by three-dimensional (3D) finite element analysis (FEA) models and gait analysis. METHODS: Two hospitalized patients with Kellgren-Lawrence grade IV varus KOA and two healthy subjects were enrolled in this study. Both patients underwent unilateral TKA. FEA models were established based on CT and MR images of the knees of the patients with KOA and healthy subjects. Gait analysis was performed using a three-dimensional motion capture system with a force plate. Three direction forces at the ankle joints were calculated by inverse dynamic analysis, which provided the load for the FEA models. The total contact forces of the knee joints were also calculated by inverse dynamic analysis to enable comparisons with the results from the FEA models. The total knee contact forces, maximum von Mises stress, and stress distribution of the medial plateau were compared between the patients and healthy subjects. The distributions of the medial plateau force at 2 and 6 months postoperatively were compared with the distributions of the forces preoperatively and those in the healthy subjects. RESULTS: During static standing, the medial plateau bore the most of the total contact forces in the knees with varus KOA (90.78% for patient 1 and 93.53% for patient 2) compared with 64.75 ± 3.34% of the total force in the healthy knees. At the first and second peaks of the ground reaction force during the stance phase of a gait cycle, the medial plateau bore a much higher percentage of contact forces in patients with KOA (74.78% and 86.48%, respectively, for patient 1; 70.68% and 83.56%, respectively, for patient 2) than healthy subjects (61.06% ± 3.43% at the first peak and 72.09% ± 1.83% at the second peak). Two months after TKA, the percentages of contact forces on the medial tibial plateau were 79.65%-85.19% at the first and second peaks of ground reaction forces during the stance phase of a gait cycle, and the percentages decreased to 53.99% - 68.13% 6 months after TKA. CONCLUSION: FEA showed that TKA effectively restored the distribution of tibio-femoral contact forces during static standing and walking, especially 6 months after the surgery. The changes in the gait were consistent with the changes in the contact force distribution calculated by the FEA model.

Multibody Computer Model of the Entire Equine Forelimb Simulates Forces Causing Catastrophic Fractures of the Carpus during a Traditional Race.

A catastrophic fracture of the radial carpal bone experienced by a racehorse during a Palio race was analyzed. Computational modelling of the carpal joint at the point of failure informed by live data was generated using a multibody code for dynamics simulation. The circuit design in a turn, the speed of the animal and the surface characteristics were considered in the model. A macroscopic examination of the cartilage, micro-CT and histology were performed on the radio-carpal joint of the limb that sustained the fracture. The model predicted the points of contact forces generated at the level of the radio-carpal joint where the fracture occurred. Articular surfaces of the distal radius, together with the proximal articular surface of small carpal bones, exhibited diffuse wear lines, erosions of the articular cartilage and subchondral bone exposure. Even though the data in this study originated from a single fracture and further work will be required to validate this approach, this study highlights the potential correlation between elevated impact forces generated at the level of contact surfaces of the carpal joint during a turn and cartilage breakdown in the absence of pre-existing pathology. Computer modelling resulted in a useful tool to inversely calculate internal forces generated during specific conditions that cannot be reproduced in-vivo because of ethical concerns.

Estimation of Spinal Loading During Manual Materials Handling Using Inertial Motion Capture.

Musculoskeletal models have traditionally relied on measurements of segment kinematics and ground reaction forces and moments (GRF&Ms) from marked-based motion capture and floor-mounted force plates, which are typically limited to laboratory settings. Recent advances in inertial motion capture (IMC) as well as methods for predicting GRF&Ms have enabled the acquisition of these input data in the field. Therefore, this study evaluated the concurrent validity of a novel methodology for estimating the dynamic loading of the lumbar spine during manual materials handling based on a musculoskeletal model driven exclusively using IMC data and predicted GRF&Ms. Trunk kinematics, GRF&Ms, L4-L5 joint reaction forces (JRFs) and erector spinae muscle forces from 13 subjects performing various lifting and transferring tasks were compared to a model driven by simultaneously recorded skin-marker trajectories and force plate data. Moderate to excellent correlations and relatively low magnitude differences were found for the L4-L5 axial compression, erector spinae muscle and vertical ground reaction forces during symmetrical and asymmetrical lifting, but discrepancies were also identified between the models, particularly for the trunk kinematics and L4-L5 shear forces. Based on these results, the presented methodology can be applied for estimating the relative L4-L5 axial compression forces under dynamic conditions during manual materials handling in the field.

The effect of saddle height and saddle position changes from pedal on muscles and joints behaviors in ergometer: A parametric study.

The physical activities such as pedaling can affect the lower limb muscles strength and rehabilitation. Improper pedaling can cause injury. In this study, we would investigate the effects of saddle place (saddle position and saddle height) on the behavior of muscles and joints. Moreover, we would try to reveal the relationship between the muscles activity (Act) and the joints reaction forces (F) and saddle position and saddle height. To this end, the pedaling conditions are obtained from the biomechanical model of the human movement system presented in AnyBody software. The variations in 12 muscles Act and total, normal and shear F of ankle, knee and hip joints are studied for the various saddle places in the pedaling feasible range. The relationships of those muscles Act and joints F are predicted by the response surface method. The results indicate that the muscles and the joints behavior changes for various saddle position and saddle height. The maximum and the minimum of the total response are acquired in the ankle and hip joints, respectively. In contrast to the ankle and hip joints, the knee shear response is greater than the normal response. The predictive models of the muscles Act and the joints F (the regression coefficients (R2) are 0.60-0.95 and 0.76-0.97, respectively) indicate their nonlinear behavior with saddle position and saddle height variations. Studying the muscles and joints behavior in different pedaling condition can be helpful for the suitable saddle placement in order for rehabilitation, muscles soreness reduction, and joints disorder treatment.

Inverse dynamic modelling of jumping in the red-legged running frog, Kassina maculata.

Although the red-legged running frog, Kassina maculata, is secondarily a walker/runner, it retains the capacity for multiple locomotor modes, including jumping at a wide range of angles (nearly 70 deg). Using simultaneous hind limb kinematics and single-foot ground reaction forces, we performed inverse dynamics analyses to calculate moment arms and torques about the hind limb joints during jumping at different angles in K. maculata. We show that forward thrust is generated primarily at the hip and ankle, while body elevation is primarily driven by the ankle. Steeper jumps are achieved by increased thrust at the hip and ankle and greater downward rotation of the distal limb segments. Because of its proximity to the GRF vector, knee posture appears to be important in controlling torque directions about this joint and, potentially, torque magnitudes at more distal joints. Other factors correlated with higher jump angles include increased body angle in the preparatory phase, faster joint openings and increased joint excursion, higher ventrally directed force, and greater acceleration and velocity. Finally, we demonstrate that jumping performance in K. maculata does not appear to be compromised by presumed adaptation to walking/running. Our results provide new insights into how frogs engage in a wide range of locomotor behaviours and the multi-functionality of anuran limbs.

Experimental study on the half-squat parachute landing for driving and validating the inverse dynamic simulation / 医用生物力学

Objective To provide data for establishing, driving and validating the inverse dynamics model of AnyBody Modeling System, the simulated half squat parachute landing experiment was designed and relevant data were collected. Method The subject was required to jump from a 0.32 m high platform to simulate the half squat parachute landing. The kinematic parameter of lower extremity joint, the ground reaction force and the surface electromyogram (SEMG) of four main muscles in the lower extremity joint were measured simultaneously. Results The angle changes of hip, knee and ankle along with time in three anatomical planes, the ground reaction force of right foot and the trajectory of the center of pressure were collected within 1 second just before and after the subject landing. These data would be used to drive the muscleskeletal model, while the data for measuring electromyogram activity would be used to validate the model. Conclusions The experiment meets the requirement of muscleskeletal model analysis, which can be used for further study of half squat parachute landing.