Studentenprojekte
Developing image registration and segmentation for multi-contrast quantitative MRI
This project aims to develop and improve methods for image registration and segmentation in quantitative MRI.
Schlagwörter
Cardiovascular magnetic resonance, image registration, image segmentation
Labels
Semester Project , Master Thesis
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-12-01 , Frühester Start: 2026-02-01
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Vousten Vincent
Themen Engineering and Technology
Generation of Physics-Based Synthetic Cerebrovascular 4D Flow MRI Data
Time-resolved volumetric phase-contrast magnetic resonance imaging (4D flow MRI) offers a non-invasive method to capture in-vivo blood flow patterns in the brain. However the accuracy and precision of the measurements cannot be quantified due to the lack of paired ground truth flow data with cerebrovascular MR measurements.
Schlagwörter
4D Flow MRI, computational fluid dynamics, modeling, blood flow, meshing
Labels
Bachelor Thesis
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-12-01 , Frühester Start: 2025-12-01 , Spätestes Ende: 2026-08-01
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Dirix Pietro
Themen Engineering and Technology
Deep Learning-based Cardiac Diffusion Tensor Imaging Denoising and Tensor Fitting
The project aims to investigate the potential of deep learning methods for image denoising and diffusion tensor fitting in cardiac diffusion tensor imaging (cDTI). By mitigating the modality’s inherently low SNR, deep learning approaches may enable shorter scan times and facilitate the translation of cDTI from a research tool into routine clinical practice.
Schlagwörter
deep learning, cardiac diffusion tensor imaging
Labels
Semester Project , Master Thesis
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-12-01 , Frühester Start: 2026-02-01 , Spätestes Ende: 2026-09-30
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Haltmeier Sandra
Themen Engineering and Technology
Evaluating nnU-Net Robustness for Aortic and Ventricular Segmentation in heavily undersampled 3D Radial PC-bSSFP
This project evaluates how pretrained nnU-Net models perform when input images are progressively degraded to mimic the undersampling and artifacts of 3D radial PC-bSSFP, determining up to what degradation level reliable aortic and ventricular segmentation—and resulting physiological parameters—can still be obtained.
Schlagwörter
3D radial PC-bSSFP, Segmentation, nnU-Net, Deep learning, Aortic flow, Undersampling simulation
Labels
Semester Project
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-12-01 , Frühester Start: 2026-03-01 , Spätestes Ende: 2026-09-30
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Malich Jacob
Themen Engineering and Technology
Forecasting Trigger Point for Gated MRI Acquisition
The project focuses on applying and evaluating methods to forecast the next physiological gating event in time-resolved MRI using past cardiac and respiratory signals.
Schlagwörter
Dynamic MRI, Cardiac Gating, Signal Processing, Machine Learning
Labels
Semester Project , Bachelor Thesis , Master Thesis
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-12-01 , Frühester Start: 2026-02-16 , Spätestes Ende: 2026-07-31
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Emery Sébastien
Themen Information, Computing and Communication Sciences , Engineering and Technology
Using (Pre-)Balancing Gradients in 3D Radial PC-bSSFP to extract Multi-Echo Water–Fat Separated Images
This project investigates whether half-echo signals generated by pre-/rephasing gradients in 3D radial PC-bSSFP can be exploited to obtain multi-echo data for water–fat separation.
Schlagwörter
Water–fat imaging, 3D radial PC-bSSFP, Multi-echo MRI, IDEAL, 4D Flow MRI, Low-field MRI, MAP Image reconstruction, Half-echo acquisitions
Labels
Semester Project
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-12-01 , Frühester Start: 2026-03-01 , Spätestes Ende: 2026-11-30
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Malich Jacob
Themen Engineering and Technology
Physics-informed deep learning for super-resolution of 4D Flow MRI
The project focuses on applying and evaluating an existing physics-informed deep learning-based super-resolution using synthetic aortic 4D flow MRI
Schlagwörter
4D flow MRI, deep learning, medical image analysis, neural networks, physics-informed, artificial intelligence, super-resolution
Labels
Semester Project , Bachelor Thesis
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-11-30 , Frühester Start: 2026-01-05
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Jacobs Luuk
Themen Engineering and Technology
Learning Based Estimation of Pulse Wave Velocity from Noisy Low Resolution Data
Aortic pulse wave velocity (PWV) is a key biomarker of cardiovascular health. It can be estimated using phase-contrast (PC) MRI, with both data-driven and physics-informed neural networks, but the accuracy of these methods needs to be validated against ground truth.
Schlagwörter
reduced-order modelling, neural networks, aortic flow, pulse wave velocity, vessel compliance, physics-informed
Labels
Semester Project , ETH Zurich (ETHZ)
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-11-28 , Frühester Start: 2025-12-01 , Spätestes Ende: 2026-09-01
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Dirix Pietro
Themen Engineering and Technology
Simulating Myocardial Scar Patterns for Synthetic Cardiac MRI Generation
This project aims to generate realistic myocardial scar patterns for the synthesis of LGE CMR images, with controllable parameters such as underlying pathology and patient-specific characteristics. The student will develop a controlled generation pipeline using either computational modelling techniques or machine-learning–based approaches.
Labels
Semester Project , ETH Zurich (ETHZ)
Description
Contact Details
Mehr Informationen
Publiziert seit: 2025-11-24 , Frühester Start: 2026-02-01 , Spätestes Ende: 2026-08-31
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Margolis Isabel
Themen Medical and Health Sciences , Information, Computing and Communication Sciences , Engineering and Technology
Comparing bSSFP and GRE cine MRI for automated aortic segmentation using deep learning
The aim of this project is to quantitatively compare two image sequences commonly used in cardiac MRI by employing automated segmentation based on convolutional neural networks.
Schlagwörter
Aortic segmentation, Cardiac MRI, Deep learning, Swiss Heart Study, Medical image analysis
Labels
Semester Project
PLEASE LOG IN TO SEE DESCRIPTION
Mehr Informationen
Publiziert seit: 2025-11-17 , Frühester Start: 2025-12-01 , Spätestes Ende: 2026-06-30
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Wolkerstorfer Gloria
Themen Information, Computing and Communication Sciences , Engineering and Technology
Deep learning segmentation of the atria from cardiac MRI using the Swiss Heart Study dataset
The project focuses on training deep learning networks for the segmentation of atria from cardiac MRI.
Schlagwörter
Atrial segmentation, Cardiac MRI, Deep learning, Swiss Heart Study, Medical image analysis
Labels
Semester Project , Bachelor Thesis , ETH Zurich (ETHZ)
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-11-14 , Frühester Start: 2025-12-01
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Buoso Stefano
Themen Engineering and Technology
Quantification of myocardial blood perfusion from cardiac perfusion MRI using advanced learning algorithms
Ischemia is a less-than-normal amount of blood flow to part of your body. It can happen particularly in heart and brain which causes severe life-threatening conditions. However, most of the current imaging technics only provides qualitative assessment, resulting in uncertainty in determining the severity of the disease. This study focuses on assessment of ischemia in heart muscles (myocardium) using cardiac perfusion magnetic resonance imaging (Perfusion MRI) and advanced learning based methods to quantify the level of perfusion in different locations of myocardium.
Schlagwörter
Image quantification, advanced learning algorithms, cardiac MRI
Labels
Semester Project
PLEASE LOG IN TO SEE DESCRIPTION
Mehr Informationen
Publiziert seit: 2025-11-05 , Frühester Start: 2026-03-01 , Spätestes Ende: 2026-11-30
Bewerbungen eingeschränkt auf ETH Zurich
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Yan Chang
Themen Information, Computing and Communication Sciences , Engineering and Technology
In-silico cardiac and cardiovascular modelling with physics informed neural networks
The aim of the project is to investigate the benefits, requirements and drawbacks of physics informed neural networks in the context of personalised cardiac and cardiovascular models
Schlagwörter
cardiac modelling, neural network, in-silico models, personalized medicine, reduced-order modelling, fluid dynamics, continuum mechanics, aortic flow
Labels
Master Thesis
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-11-04 , Frühester Start: 2020-09-01
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Buoso Stefano , Kozerke Sebastian, Prof
Themen Information, Computing and Communication Sciences , Engineering and Technology
Generation of synthetic cardiac phantoms for healthy and pathological anatomy and function using generative AI
The project focuses exploiting generative AI to build synthetic numerical phantom for cardiac anatomy and function suitable for representing population variability.
Schlagwörter
Generative models, deep learning, phantoms, variational autoencoders, cardiac mechanics, cardiac function, simulation
Labels
Master Thesis
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-11-04 , Frühester Start: 2023-07-01
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Buoso Stefano , Kozerke Sebastian, Prof
Themen Information, Computing and Communication Sciences , Engineering and Technology
Digital twinning with physics-informed graph neural networks
The aim of this project is to develop an approach based on physics-based graph neural networks to generate digital twins from PC-MRI data.
Schlagwörter
Physics informed neural networks, graph neural networks, digital twins, blood flow, brain, circulatory system, AI, biophysical
Labels
Master Thesis
Description
Goal
Contact Details
Mehr Informationen
Publiziert seit: 2025-11-04 , Frühester Start: 2025-11-01
Organisation(en) Cardiovascular Magnetic Resonance
Host(s) Buoso Stefano , Kozerke Sebastian, Prof
Themen Information, Computing and Communication Sciences , Engineering and Technology