Surrogate modeling of deformable joint contact using artificial neural networks

Integrating musculoskeletal simulation and machine learning: a hybrid approach for personalized ankle-foot exoskeleton assistance strategies

Introduction: Lower limb exoskeletons have shown considerable potential in assisting human walking, particularly by reducing metabolic cost (MC), leading to a surge of interest in this field in recent years. However, owing to significant individual differences and the uncertainty of movements, challenges still exist in the personalized design and control of exoskeletons in human-robot interactions. Methods: In this study, we propose a hybrid data-driven approach that integrates musculoskeletal simulation with machine learning technology to customize personalized assistance strategies efficiently and adaptively for ankle-foot exoskeletons. First, optimal assistance strategies that can theoretically minimize MC, were derived from forward muscle-driven simulations on an open-source dataset. Then, a neural network was utilized to explore the relationships among different individuals, movements, and optimal strategies, thus developing a predictive model. Results: With respect to transfer learning, our approach exhibited effectiveness and adaptability when faced with new individuals and movements. The simulation results further indicated that our approach successfully reduced the MC of calf muscles by approximately 20% compared to normal walking conditions. Discussion: This hybrid approach offers an alternative for personalizing assistance strategy that may further guide exoskeleton design.

Robotics in Arthroplasty: Historical Progression, Contemporary Applications, and Future Horizons With Artificial Intelligence (AI) Integration

Robotic technology is increasingly utilized in surgical procedures to enhance precision, particularly in tasks demanding delicate maneuvers beyond human capabilities. Robotic orthopedic surgery emerges as a dynamic and compelling technology reshaping the landscape of surgical practice. This aids surgeons in achieving enhanced accuracy and reproducibility, ultimately aiming for improved patient outcomes. As of now, the majority of these systems are in a developed stage and are gradually gaining broader adoption. These systems have to show that they are user-friendly, are successful in clinical settings, and have a good cost-effectiveness ratio before they can be widely adopted in the field of surgery. In this review, we examine the evolution of robotics in orthopedic surgery, assess its current applications, and provide insights into the future trajectory of this technology, particularly in light of advances in artificial intelligence (AI) and machine learning (ML).

Supervised machine learning for the prediction of post-operative clinical outcomes of hip and knee replacements: a review

Prediction models are being increasingly used in the medical field to identify risk factors and possible outcomes. Some of these are presently being used to develop guidelines for improving clinical practice. The application of machine learning (ML), comprising a powerful set of computational tools for analysing data, has been clearly expanding in the role of predictive modelling. This paper reviews the latest developments of supervised ML techniques that have been used to analyse data related to post-operative total hip and knee replacements. The aim was to review the most recent findings of relevant published studies by outlining the methodologies employed (most-widely used supervised ML techniques), data sources, domains, limitations of predictive analytics and the quality of predictions.

Assessing the behavior of a hybrid model of the knee with contact surrogate under parameter uncertainties

Modeling the knee is an important factor in increasing the quality of life of both healthy individuals and patients. Nevertheless, the intricate nature of the knee makes this problem complicated. In this study, an extension to an established planar knee joint model with Hertzian contact pairs is proposed with contact mechanics based on polynomial chaos expansion surrogate. Firstly, the finite element (FE) model is made representing a contact pair of sphere-to-plane type with two layers on both bodies, corresponding to the cartilage and the bone. Five variables corresponding to both geometry and material parameters are used to parametrize this model. Then, 128 distinct variants of the FE model are created based on a quasi-Monte Carlo sequence. This dataset is used to train and validate the surrogate. The trained surrogate is proven to have predictive capabilities with an average nRMSE of 0.2% in randomized test/train splits. When included in a model of the knee and tested under parameter uncertainties in Monte Carlo simulations, it results in nRMSE of 58% for angular coordinate compared to the original model with Hertzian pair. This signifies the high influence of contact formulation on the model output and the need for more physically based models in knee contact modeling.

Artificial intelligence applications in bone fractures: A bibliometric and science mapping analysis

Background

Bone fractures are a common medical issue worldwide, causing a serious economic burden on society. In recent years, the application of artificial intelligence (AI) in the field of fracture has developed rapidly, especially in fracture diagnosis, where AI has shown significant capabilities comparable to those of professional orthopedic surgeons. This study aimed to review the development process and applications of AI in the field of fracture using bibliometric analysis, while analyzing the research hotspots and future trends in the field.

Materials and methods

Studies on AI and fracture were retrieved from the Web of Science Core Collections since 1990, a retrospective bibliometric and visualized study of the filtered data was conducted through CiteSpace and Bibliometrix R package.

Results

A total of 1063 publications were included in the analysis, with the annual publication rapidly growing since 2017. China had the most publications, and the United States had the most citations. Technical University of Munich, Germany, had the most publications. Doornberg JN was the most productive author. Most research in this field was published in Scientific Reports. Doi K's 2007 review in Computerized Medical Imaging and Graphics was the most influential paper.

Conclusion

AI application in fracture has achieved outstanding results and will continue to progress. In this study, we used a bibliometric analysis to assist researchers in understanding the basic knowledge structure, research hotspots, and future trends in this field, to further promote the development of AI applications in fracture.

Applications of artificial intelligence in orthopaedic surgery

The practice of medicine is rapidly transforming as a result of technological breakthroughs. Artificial intelligence (AI) systems are becoming more and more relevant in medicine and orthopaedic surgery as a result of the nearly exponential growth in computer processing power, cloud based computing, and development, and refining of medical-task specific software algorithms. Because of the extensive role of technologies such as medical imaging that bring high sensitivity, specificity, and positive/negative prognostic value to management of orthopaedic disorders, the field is particularly ripe for the application of machine-based integration of imaging studies, among other applications. Through this review, we seek to promote awareness in the orthopaedics community of the current accomplishments and projected uses of AI and ML as described in the literature. We summarize the current state of the art in the use of ML and AI in five key orthopaedic disciplines: joint reconstruction, spine, orthopaedic oncology, trauma, and sports medicine.

Optimal Conformity Design of Tibial Insert Component Based on ISO Standard Wear Test Using Finite Element Analysis and Surrogate Model

Total knee replacement is a standard surgical treatment used to treat osteoarthritis in the knee. The implant is complicated, requiring expensive designs and testing as well as a surgical intervention. This research proposes a technique concerning the optimal conformity design of the symmetric polyethylene tibial insert component for fixed-bearing total knee arthroplasty. The Latin Hypercube Sampling (LHS) design of the experiment was used to create 30 cases of the varied tibial insert conformity that influenced the total knee replacement wear volume. The combination of finite element analysis and a surrogate model was performed to predict wear volume according to the standard of ISO-14243:2014 wear test and to determine the optimal conformity. In the first step, the results could predict wear volume between 5.50 to 72.92 mm3/106 cycle. The Kriging method of a surrogate model has then created the increased design based on the efficient global optimization (EGO) method with improving data 10 design points. The result revealed that the optimum design of tibial insert conformity in a coronal and sagittal plane was 0.70 and 0.59, respectively, with a minimizing wear volume of 3.07 mm3/106 cycle. The verification results revealed that the area surface scrape and wear volume are similar to those predicted by the experiment. The wear behavior on the tibial insert surface was asymmetry of both sides. From this study it can be concluded that the optimal conformity design of the tibial insert component can be by using a finite element and surrogate model combined with the design of conformity to the minimized wear volume.

A Conceptual Blueprint for Making Neuromusculoskeletal Models Clinically Useful

The ultimate goal of most neuromusculoskeletal modeling research is to improve the treatment of movement impairments. However, even though neuromusculoskeletal models have become more realistic anatomically, physiologically, and neurologically over the past 25 years, they have yet to make a positive impact on the design of clinical treatments for movement impairments. Such impairments are caused by common conditions such as stroke, osteoarthritis, Parkinson’s disease, spinal cord injury, cerebral palsy, limb amputation, and even cancer. The lack of clinical impact is somewhat surprising given that comparable computational technology has transformed the design of airplanes, automobiles, and other commercial products over the same time period. This paper provides the author’s personal perspective for how neuromusculoskeletal models can become clinically useful. First, the paper motivates the potential value of neuromusculoskeletal models for clinical treatment design. Next, it highlights five challenges to achieving clinical utility and provides suggestions for how to overcome them. After that, it describes clinical, technical, collaboration, and practical needs that must be addressed for neuromusculoskeletal models to fulfill their clinical potential, along with recommendations for meeting them. Finally, it discusses how more complex modeling and experimental methods could enhance neuromusculoskeletal model fidelity, personalization, and utilization. The author hopes that these ideas will provide a conceptual blueprint that will help the neuromusculoskeletal modeling research community work toward clinical utility.

Artificial Intelligence and Machine Learning in Lower Extremity Arthroplasty: A Review

BACKGROUND

Driven by the rapid development of big data and processing power, artificial intelligence and machine learning (ML) applications are poised to expand orthopedic surgery frontiers. Lower extremity arthroplasty is uniquely positioned to most dramatically benefit from ML applications given its central role in alternative payment models and the value equation.

METHODS

In this report, we discuss the origins and model specifics behind machine learning, consider its progression into healthcare, and present some of its most recent advances and applications in arthroplasty.

RESULTS

A narrative review of artificial intelligence and ML developments is summarized with specific applications to lower extremity arthroplasty, with specific lessons learned from osteoarthritis gait models, joint-specific imaging analysis, and value-based payment models.

CONCLUSION

The advancement and employment of ML provides an opportunity to provide data-driven, high performance medicine that can rapidly improve the science, economics, and delivery of lower extremity arthroplasty.

A novel training-free method for real-time prediction of femoral strain

Surrogate methods for rapid calculation of femoral strain are limited by the scope of the training data. We compared a newly developed training-free method based on the superposition principle (Superposition Principle Method, SPM) and popular surrogate methods for calculating femoral strain during activity. Finite-element calculations of femoral strain, muscle, and joint forces for five different activity types were obtained previously. Multi-linear regression, multivariate adaptive regression splines, and Gaussian process were trained for 50, 100, 200, and 300 random samples generated using Latin Hypercube (LH) and Design of Experiment (DOE) sampling. The SPM method used weighted linear combinations of 173 activity-independent finite-element analyses accounting for each muscle and hip contact force. Across the surrogate methods, we found that 200 DOE samples consistently provided low error (RMSE < 100 µε), with model construction time ranging from 3.8 to 63.3 h and prediction time ranging from 6 to 1236 s per activity. The SPM method provided the lowest error (RMSE = 40 µε), the fastest model construction time (3.2 h) and the second fastest prediction time per activity (36 s) after Multi-linear Regression (6 s). The SPM method will enable large numerical studies of femoral strain and will narrow the gap between bone strain prediction and real-time clinical applications.

Deep Learning for Musculoskeletal Force Prediction

Musculoskeletal models permit the determination of internal forces acting during dynamic movement, which is clinically useful, but traditional methods may suffer from slowness and a need for extensive input data. Recently, there has been interest in the use of supervised learning to build approximate models for computationally demanding processes, with benefits in speed and flexibility. Here, we use a deep neural network to learn the mapping from movement space to muscle space. Trained on a set of kinematic, kinetic and electromyographic measurements from 156 subjects during gait, the network's predictions of internal force magnitudes show good concordance with those derived by musculoskeletal modelling. In a separate set of experiments, training on data from the most widely known benchmarks of modelling performance, the international Grand Challenge competitions, generates predictions that better those of the winning submissions in four of the six competitions. Computational speedup facilitates incorporation into a lab-based system permitting real-time estimation of forces, and interrogation of the trained neural networks provides novel insights into population-level relationships between kinematic and kinetic factors.

Efficacy and efficiency of multivariate linear regression for rapid prediction of femoral strain fields during activity

Multivariate Linear Regression-based (MLR) surrogate models were explored to reduce the computational cost of predicting femoral strains during normal activity in comparison with finite element analysis. The musculoskeletal model of one individual, the finite-element model of the right femur, and experimental force and motion data for normal walking, fast walking, stair ascent, stair descent, and rising from a chair were obtained from a previous study. Equivalent Von Mises strain was calculated for 1000 frames uniformly distributed across activities. MLR surrogate models were generated using training sets of 50, 100, 200 and 300 samples. The finite-element and MLR analyses were compared using linear regression. The Root Mean Square Error (RMSE) and the 95th percentile of the strain error distribution were used as indicators of average and peak error. The MLR model trained using 200 samples (RMSE < 108 µε; peak error < 228 µε) was used as a reference. The finite-element method required 66 s per frame on a standard desktop computer. The MLR model required 0.1 s per frame plus 1848 s of training time. RMSE ranged from 1.2% to 1.3% while peak error ranged from 2.2% to 3.6% of the maximum micro-strain (5020 µε). Performance within an activity was lower during early and late stance, with RMSE of 4.1% and peak error of 8.6% of the maximum computed micro-strain. These results show that MLR surrogate models may be used to rapidly and accurately estimate strain fields in long bones during daily physical activity.

Simulating contact using the elastic foundation algorithm in OpenSim

Modeling joint contact in OpenSim is not well understood. This study systematically investigated the variables associated with the elastic foundation contact model within OpenSim by performing a series of controlled benchtop experiment and concomitant simulations. Four metal-on-plastic interactions were modeled, including a model of a total knee replacement (TKR). Load-displacement curves were recorded during cyclic loading between 100 and 750 N. Geometries were imported and into OpenSim and contact mechanics were modeled with the on-board elastic foundation algorithm. A hybrid optimization algorithm determined that stiffness and dissipation coefficients for TKR implants were 1.52 × 10¹⁰ N/m and 57.7 Ns/m, respectively. Estimations of contact forces were 10.2% of blinded experimental data (average root mean square error: 76.82 ± 11.47 N). In the second portion of this study, freely available eTibia TKR renderings were used to test the ubiquity of the tuning parameters. They were also used to perform a sensitivity analysis of material stiffness and mesh density with regard to penetration depth and computational time. When a stiffness of 1 × 10¹⁰ was applied to an eTibia model with 5000 faces, a 100 kg load caused 0.259 mm of penetration. Under the same conditions, the tuned model experienced 0.300 mm of penetration. Material stiffnesses between 1 × 10¹³ and 1 × 10¹⁵ N/m increased computation time by factors of 12-23. This study provides much needed clarity regarding the use of the OpenSim EF algorithm. It also demonstrates the utility of OpenSim to model deformable materials and complex geometries, and this approach can be adapted to make reasonable estimations for both natural and surgically modified joints.

A computational framework for simultaneous estimation of muscle and joint contact forces and body motion using optimization and surrogate modeling

Concurrent estimation of muscle activations, joint contact forces, and joint kinematics by means of gradient-based optimization of musculoskeletal models is hindered by computationally expensive and non-smooth joint contact and muscle wrapping algorithms. We present a framework that simultaneously speeds up computation and removes sources of non-smoothness from muscle force optimizations using a combination of parallelization and surrogate modeling, with special emphasis on a novel method for modeling joint contact as a surrogate model of a static analysis. The approach allows one to efficiently introduce elastic joint contact models within static and dynamic optimizations of human motion. We demonstrate the approach by performing two optimizations, one static and one dynamic, using a pelvis-leg musculoskeletal model undergoing a gait cycle. We observed convergence on the order of seconds for a static optimization time frame and on the order of minutes for an entire dynamic optimization. The presented framework may facilitate model-based efforts to predict how planned surgical or rehabilitation interventions will affect post-treatment joint and muscle function.

Evaluation of a Surrogate Contact Model in Force-Dependent Kinematic Simulations of Total Knee Replacement

Knowing the forces in the human body is of great clinical interest and musculoskeletal (MS) models are the most commonly used tool to estimate them in vivo. Unfortunately, the process of computing muscle, joint contact, and ligament forces simultaneously is computationally highly demanding. The goal of this study was to develop a fast surrogate model of the tibiofemoral (TF) contact in a total knee replacement (TKR) model and apply it to force-dependent kinematic (FDK) simulations of activities of daily living (ADLs). Multiple domains were populated with sample points from the reference TKR contact model, based on reference simulations and design-of-experiments. Artificial neural networks (ANN) learned the relationship between TF pose and loads from the medial and lateral sides of the TKR implant. Normal and right-turn gait, rising-from-a-chair, and a squat were simulated using both surrogate and reference contact models. Compared to the reference contact model, the surrogate contact model predicted TF forces with a root-mean-square error (RMSE) lower than 10 N and TF moments lower than 0.3 N·m over all simulated activities. Secondary knee kinematics were predicted with RMSE lower than 0.2 mm and 0.2 deg. Simulations that used the surrogate contact model ran on average three times faster than those using the reference model, allowing the simulation of a full gait cycle in 4.5 min. This modeling approach proved fast and accurate enough to perform extensive parametric analyses, such as simulating subject-specific variations and surgical-related factors in TKR.

Real-Time Prediction of Temperature Elevation During Robotic Bone Drilling Using the Torque Signal

Bone drilling is a surgical procedure commonly required in many surgical fields, particularly orthopedics, dentistry and head and neck surgeries. While the long-term effects of thermal bone necrosis are unknown, the thermal damage to nerves in spinal or otolaryngological surgeries might lead to partial paralysis. Previous models to predict the temperature elevation have been suggested, but were not validated or have the disadvantages of computation time and complexity which does not allow real time predictions. Within this study, an analytical temperature prediction model is proposed which uses the torque signal of the drilling process to model the heat production of the drill bit. A simple Green's disk source function is used to solve the three dimensional heat equation along the drilling axis. Additionally, an extensive experimental study was carried out to validate the model. A custom CNC-setup with a load cell and a thermal camera was used to measure the axial drilling torque and force as well as temperature elevations. Bones with different sets of bone volume fraction were drilled with two drill bits ([Formula: see text]1.8 mm and [Formula: see text]2.5 mm) and repeated eight times. The model was calibrated with 5 of 40 measurements and successfully validated with the rest of the data ([Formula: see text]C). It was also found that the temperature elevation can be predicted using only the torque signal of the drilling process. In the future, the model could be used to monitor and control the drilling process of surgeries close to vulnerable structures.

A Computational Efficient Method to Assess the Sensitivity of Finite-Element Models: An Illustration With the Hemipelvis

Assessing the sensitivity of a finite-element (FE) model to uncertainties in geometric parameters and material properties is a fundamental step in understanding the reliability of model predictions. However, the computational cost of individual simulations and the large number of required models limits comprehensive quantification of model sensitivity. To quickly assess the sensitivity of an FE model, we built linear and Kriging surrogate models of an FE model of the intact hemipelvis. The percentage of the total sum of squares (%TSS) was used to determine the most influential input parameters and their possible interactions on the median, 95th percentile and maximum equivalent strains. We assessed the surrogate models by comparing their predictions to those of a full factorial design of FE simulations. The Kriging surrogate model accurately predicted all output metrics based on a training set of 30 analyses (R2 = 0.99). There was good agreement between the Kriging surrogate model and the full factorial design in determining the most influential input parameters and interactions. For the median, 95th percentile and maximum equivalent strain, the bone geometry (60%, 52%, and 76%, respectively) was the most influential input parameter. The interactions between bone geometry and cancellous bone modulus (13%) and bone geometry and cortical bone thickness (7%) were also influential terms on the output metrics. This study demonstrates a method with a low time and computational cost to quantify the sensitivity of an FE model. It can be applied to FE models in computational orthopaedic biomechanics in order to understand the reliability of predictions.

An Open-Source Toolbox for Surrogate Modeling of Joint Contact Mechanics

GOAL

Incorporation of elastic joint contact models into simulations of human movement could facilitate studying the interactions between muscles, ligaments, and bones. Unfortunately, elastic joint contact models are often too expensive computationally to be used within iterative simulation frameworks. This limitation can be overcome by using fast and accurate surrogate contact models that fit or interpolate input-output data sampled from existing elastic contact models. However, construction of surrogate contact models remains an arduous task. The aim of this paper is to introduce an open-source program called Surrogate Contact Modeling Toolbox (SCMT) that facilitates surrogate contact model creation, evaluation, and use.

METHODS

SCMT interacts with the third-party software FEBio to perform elastic contact analyses of finite-element models and uses MATLAB to train neural networks that fit the input-output contact data. SCMT features sample point generation for multiple domains, automated sampling, sample point filtering, and surrogate model training and testing.

RESULTS

An overview of the software is presented along with two example applications. The first example demonstrates creation of surrogate contact models of artificial tibiofemoral and patellofemoral joints and evaluates their computational speed and accuracy, while the second demonstrates the use of surrogate contact models in a forward dynamic simulation of an open-chain leg extension-flexion motion.

CONCLUSION

SCMT facilitates the creation of computationally fast and accurate surrogate contact models. Additionally, it serves as a bridge between FEBio and OpenSim musculoskeletal modeling software.

SIGNIFICANCE

Researchers may now create and deploy surrogate models of elastic joint contact with minimal effort.