Physics Driven Neural Network and Its Application in Fast Cardiovascular Disease Diagnosis
Project Description
This project learns how to use PINN to solve CFD problem in the cardiovascular system for various diagnosis and treatment purposes.
The PINN approach can combine both physics (explainable) and data-driven approach (Fast). It enables us to develop a reliable AI tool for future clinical tool application.
The PINN approach can combine both physics (explainable) and data-driven approach (Fast). It enables us to develop a reliable AI tool for future clinical tool application.
Supervisor
HU, Wenqi
Quota
2
Course type
UROP1000
UROP1100
UROP2100
UROP3100
UROP3200
UROP4100
Applicant's Roles
Learrn the PINN tech we have.
Develop this tech.
Apply it to real medical imaging to check its performance on diagnosis.
Develop this tech.
Apply it to real medical imaging to check its performance on diagnosis.
Applicant's Learning Objectives
Learn the PINN approach
Learn the basics about hemodynamics
Learn the biomechanics of cardiovascular disease
Learn the basics about hemodynamics
Learn the biomechanics of cardiovascular disease
Complexity of the project
Moderate