Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Computational Fluid Dynamics (CFD) Simulation for Evaluation of Aortic Blood Flow, Machine Learning
Set up of a CFD simulation of a human aortic arch in order to evaluate current patterns using machine learning.
Keywords: CFD, computational fluid dynamics, machine learning, open source, Magnetic Resonance Imaging, 4D Flow MRI, Python, C
Computational Fluid Dynamics (CFD) provides a prediction of fluid flows by means of
- mathematical modelling (partial differential equations)
- numerical methods (discretization and solution techniques)
- software tools (solvers, pre- and postprocessing utilities)
The fundamental basis of most CFD problems are the Navier-Stokes equations, which are solved using numerical methods.
4D Flow MRI enables us to measure blood flow velocities in the human body in three dimensions. However, in order to keep measurement time in an clinically acceptable time range, undersampling is performed and outputs a distorted image. A CFD model providing a super resolution "ground truth" acts as an important sandbox example to try out new reconstruction and machine learning algorithms.
At the Institute for Biomedical Engineering we investigate how modern machine learning algorithms can benefit the reconstruction of 4D Flow MR images. Since the training (measured) data is limited, CFD simulation output would be of great use. This work comprises CFD simulations in OpenFoam and Machine Learning in TensorFlow.
Side effects include:
- The student will get to know the "de-facto" standard in scientific CFD simulations, namely OpenFoam.
- The student will get access to our HPC at IBT to perform computations.
- The student will have the opportunity to actively contribute to current research projects that are ongoing at IBT.
Computational Fluid Dynamics (CFD) provides a prediction of fluid flows by means of
- numerical methods (discretization and solution techniques)
- software tools (solvers, pre- and postprocessing utilities)
The fundamental basis of most CFD problems are the Navier-Stokes equations, which are solved using numerical methods.
4D Flow MRI enables us to measure blood flow velocities in the human body in three dimensions. However, in order to keep measurement time in an clinically acceptable time range, undersampling is performed and outputs a distorted image. A CFD model providing a super resolution "ground truth" acts as an important sandbox example to try out new reconstruction and machine learning algorithms.
At the Institute for Biomedical Engineering we investigate how modern machine learning algorithms can benefit the reconstruction of 4D Flow MR images. Since the training (measured) data is limited, CFD simulation output would be of great use. This work comprises CFD simulations in OpenFoam and Machine Learning in TensorFlow.
Side effects include:
- The student will get to know the "de-facto" standard in scientific CFD simulations, namely OpenFoam.
- The student will get access to our HPC at IBT to perform computations.
- The student will have the opportunity to actively contribute to current research projects that are ongoing at IBT.
The project extent can be tailored to the interest and background of the student. Main topics of this ad are
- CFD simulation with Open Source tools that are widely used (FreeCAD, OpenFoam, Paraview)
- Using machine learning algorithms for evaluation of current patterns in the aortic arch (TensorFlow)
Please contact dillinger@biomed.ee.ethz.ch if you are interested in the topic.
Interested students are asked to send a CV and a transcript of records.
For students from abroad, we also need to know how they plan to organize funding for their stay.
The project extent can be tailored to the interest and background of the student. Main topics of this ad are
- CFD simulation with Open Source tools that are widely used (FreeCAD, OpenFoam, Paraview)
- Using machine learning algorithms for evaluation of current patterns in the aortic arch (TensorFlow)
Please contact dillinger@biomed.ee.ethz.ch if you are interested in the topic.
Interested students are asked to send a CV and a transcript of records.
For students from abroad, we also need to know how they plan to organize funding for their stay.
Supervisor: Hannes Dillinger (dillinger@biomed.ee.ethz.ch)
Professor: Prof. Dr. Sebastian Kozerke
Supervisor: Hannes Dillinger (dillinger@biomed.ee.ethz.ch)