Surrogate modeling of deformable joint contact using artificial neural networks

Abstract

Deformable joint contact models can be used to estimate loading conditions for cartilage-cartilage, implant-implant, human-orthotic, and foot-ground interactions. However, contact evaluations are often so expensive computationally that they can be prohibitive for simulations or optimizations requiring thousands or even millions of contact evaluations. To overcome this limitation, we developed a novel surrogate contact modeling method based on artificial neural networks (ANNs). The method uses special sampling techniques to gather input-output data points from an original (slow) contact model in multiple domains of input space, where each domain represents a different physical situation likely to be encountered. For each contact force and torque output by the original contact model, a multi-layer feed-forward ANN is defined, trained, and incorporated into a surrogate contact model. As an evaluation problem, we created an ANN-based surrogate contact model of an artificial tibiofemoral joint using over 75,000 evaluations of a fine-grid elastic foundation (EF) contact model. The surrogate contact model computed contact forces and torques about 1000 times faster than a less accurate coarse grid EF contact model. Furthermore, the surrogate contact model was seven times more accurate than the coarse grid EF contact model within the input domain of a walking motion. For larger input domains, the surrogate contact model showed the expected trend of increasing error with increasing domain size. In addition, the surrogate contact model was able to identify out-of-contact situations with high accuracy. Computational contact models created using our proposed ANN approach may remove an important computational bottleneck from musculoskeletal simulations or optimizations incorporating deformable joint contact models.

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

“…These models utilized a variety of techniques, such as response surface optimization [6], Lazy Learning [4], nonlinear dynamic models [5], Kriging [8][9][10], and 94 artificial neural networks (ANN) [5,7,11]. Recently, Eskinazi and Fregly (2015) proposed a surrogate modeling approach based on ANN to accelerate an FE deformable contact 96 model of TKR [11]. Artificial neural networks are known, among others, for their ability to learn virtually any complex relationship between a set of input and output variables 98…”

Toward modeling locomotion using electromyography‐informed 3D models: application to cerebral palsy

“…However, the associated large computational requirements and challenges in validating resulting predictions may represent additional factors limiting the effective translation to clinical setting. In this scenario, high‐fidelity formulations could be synthesized into simpler, surrogate models that describe the complex musculoskeletal structures using computationally efficient structures such as multidimensional splines, regressive polynomials, or artificial neural networks . These could be coupled with neural‐based EMG‐informed modeling and create fully personalized, clinically viable formulations, which capture neurophysiological, anatomical, and morphological abnormalities altogether.…”

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

“…The TF model outputs medial and lateral superior-inferior forces ( $F_y^{\mathit{med}}$ and $F_y^{\mathit{lat}}$) to describe the medial-lateral load split. The inputs and outputs to each stage were chosen using a previously defined method [35]. For both models, the contact loads that were highly sensitive to pose parameter variations were fit as functions of the pose parameters while the insensitive loads were fit as functions of the pose parameters and the sensitive loads calculated in the earlier stages.…”