Deep learning acceleration of Total Lagrangian Explicit Dynamics for soft tissue mechanics

Analytically Integratable Zero-restlength Springs for Capturing Dynamic Modes unrepresented by Quasistatic Neural Networks

“…[Chentanez et al 2020] proposes a transition-based model with position and linear/angular velocity of the body as network input (in addition to a state-based model). [Meister et al 2020] uses a fully connected network to predict node-wise acceleration for total Lagrangian explicit dynamics. [Deng et al 2020] proposes a convolutional long short-term memory (LSTM) layer to capture elastic force propagation.…”

Deep learning for model order reduction of multibody systems to minimal coordinates

“…In this work, a deep learning [12,13] approach to obtain the mapping from general to minimal coordinates in case of rigid mechanisms is proposed. Over the past years, several authors have explored the use of machine learning schemes to accelerate or improve dedicated mechanical analyses [14][15][16]. More specifically, we propose the use of a trained AutoEncoder (AE) neural network to approximate the mapping between the minimal and the system coordinates.…”

A Meshfree Method for Deformation Field Reconstruction of Soft Tissue in Needle Insertion

“…As needle insertion into an organ is a non-linear problem of continuum mechanics which involves large deformations and large strains with non-linear material, and the FE method is incompatible with the assumption of infinitesimal deformations [7]. To improve the computation accuracy, different computation models including the co-rotational finite elements [8], [9], and the Total Lagrangian explicit dynamics algorithm [10][11][12] have been proposed. Although both can be applied in real-time deformation computation, they are still based on the theory of small deformation and linear material properties.…”