NCSI: Numerical Simulation of Knee Joint Replacement using Finite Element and Surrogate Modeling

Status	Completed
Mentor Name	Jun Li
Mentor's XSEDE Affiliation	Research Allocation
Mentor Has Been in XSEDE Community	1-2 years
Project Title	Numerical Simulation of Knee Joint Replacement using Finite Element and Surrogate Modeling
Summary	Knee-joint replacement is a procedure of replacing an injured joint with an artificial one, or prosthesis to mimic the function of a knee, taking into consideration of the patient's age, weight, activity level, and overall health. This project aims to develop accurate and efficient simulations that reproduce the artificial-knee tibiofemoral kinematics, which involves: 1. Develop finite element (FE) simulators for the knee dynamic and the knee extension tests. 2. Develop neural networks (NN) based on self-organizing models and nature-inspired algorithms such as genetic programming (GP) that learn from FE simulations of the artificial-tibiofemoral kinematics.
Job Description	The student will build a virtual knee-joint simulator for the knee extension test to assess the knee-joint kinematics in finite element software. The student will collect the data from the finite element simulations for the knee joint replacement to train the NN. The student will build a self-organizing NN optimized using nature-inspired algorithms from open-source packages to replicate the finite element simulation for the knee extension test.
Computational Resources	We will start this project on a local HPC system with a reduced size model to let the student practice different self-organizing methods and genetic programming to create the digital twin of the knee-joint replacement based on finite element simulations. Then we plan to move to cloud-based XSEDE resources of Stampede 2 for a large size model. When time allows, we will further conduct performance tests of large models on Stampede 2.
Contribution to Community	This project will train undergraduate students to do computational research of finite element and neural network modeling. The project will also lead to 1) build a virtual simulator for the knee-joint replacement to mimic the knee extension test using finite element analysis; 2) findings and data collected from finite element simulations will be used to establish an efficient surrogate model using neural networks.
Position Type	Apprentice
Training Plan	Training tutorials will be provided for the student to use finite element analysis to simulate the artificial-tibiofemoral kinematics. The student will be prepared to use the self-organizing method based on polynomial neural networks and optimization tools in open source packages using available examples. In addition, the student will be trained to be familiar with Linux operating system and High-performance computing environment.
Student Prerequisites/Conditions/Qualifications	Must have finite element modeling background and programming skills such as in Matlab or Python to develop user scripts and build models.
Duration	Summer
Start Date	06/01/2021
End Date	08/15/2021