2nd November, Fall meeting 2022: Simulation and Synthesis

We are excited to meet live with all of you again. On Wednessday 2nd of November 2022 we hope to see you in Utrecht.

The NVPHBV Fall Meeting will focus on simulation and synthesis in pattern recognition and image processing R&D.

Simulation and synthesis can be beneficial to understand, develop, optimize, evaluate, and validate image acquisition, reconstruction, processing, and analysis/recognition methods. It is, for example, increasingly used in the field medical imaging, especially to generate ample data for training machine/deep learning based solutions. But it is also used in many other imaging-related fields.

Our meeting will consist of two parts:

In the morning (10:30 – lunch), we will host the final symposium of the EU Marie Curie openGTN project (www.opengtn.eu) that focused on the generation of ample magnetic resonance image (MRI) data for deep learning training. Exciting new methods of MRI simulation and synthesis and their application to image segmentation will be presented by the involved PhD researchers. Keynote presentation will be given by prof. dr. Sebastian Kozerke from ETH Zurich.

For the afternoon (after lunch – 17:30), we will have two keynote speakers( Dr. Jakub Tomczak, Vrije Universiteit Amsterdam and Dr. Cor Datema, Dynaxion ) , abstract presentations as well as a networking event. The winner of the best-abstract award will be announced at the end of the meeting.

The meeting (incl. lunch, coffee/tea and drinks) is free for members of NVPHBV, and only 15 euros for others, but free if you become a member (for only 15 euro/year)

Hourly Schedule

Fall Meeting 2022

10:00 - 10:30
Welcome & Coffee

10:30 - 10:35
Brief introduction UMCU / ISI

Speakers: Dr. Hugo Kuijf

10:35 - 10:45
Introduction to the openGTN project

Speakers: Prof. dr. Marcel Breeuwer

10:45 - 11:30
Importance, status and potential of simulation and synthesis in medical imaging

Speakers: Prof. Dr. Sebastian Kozerke

11:30 - 11:50
Cardiac MR image simulation and synthesis

Speakers: Sina Amirrajab

11:50 - 12:10
Brain MR image simulation and synthesis

Speakers: Aymen Ayaz

12:10 - 12:30
Towards more generalizable and robust MR image segmentation

Speakers: Yasmina Al Khalil

12:30 - 13:30
Lunch and posters

13:30 - 14:15
Generative Artificial Intelligence

Speakers: Dr. Jakub Tomczak

14:15 - 14:30
Facial expression manipulation for improved and personalized facial action estimation
Itir Önal Ertuğrul

14:30 - 14:45
Framework for developing a machine learning based MRI quality tool using simulated data
Bernardo Pereira

14:45 - 15:00
Equivariant wall shear stress estimation on arterial surface meshes
Julian Suk

15:00 - 15:30
Break - coffee, tea, soda’s and networking

15:30 - 16:15
Simulations supported development of a neutron-based security scanner

Speakers: Dr. Cor Datema

16:15 - 16:30
Early Warning Scan for breast cancer
Tom Sanders

16:30 - 16:45
Frechet Face Distance – A New Metric for Inpainting Facial Images
Raj Kumar Ashokan / Gayathri Dhanapal

16:45 - 17:00
Frequency Shortcut Learning in Neural Networks
Shunxin Wang

17:00 - 18:00
Reception – Poster award, drinks & networking

Dr. Hugo Kuijf

Image Sciences Institute, UMC Utrecht

Hugo Kuijf is assistant professor at the Image Sciences Institute, UMC Utrecht; programme coordinator of the MSc programme Medical Imaging; member of the Board of Examiners of the Graduate School of Life Sciences; chair of the Education Committee of the PhD programme Medical Imaging; and university lecturer at Eindhoven University of Technology. Hugo Kuijf graduated in Computer Science at Utrecht University in 2009, with academic minors in Software Engineering and Game- and Media Technology. In 2013, he received his PhD in Medical Imaging after defending his thesis entitled "Image processing techniques for quantification and assessment of brain MRI". His research focuses on innovative image processing and (deep) machine learning techniques for the quantification and assessment of brain MR images. These techniques are applied in the context of brain anatomy and pathology, in particular small vessel disease. Semi-automated techniques for the detection of microbleeds, microinfarcts, small arteries and veins, perivascular spaces, and white matter hyperintensities are developed; including lesion-symptom mapping solutions. Development and utilization of modern machine learning and deep learning techniques (also known as "artificial intelligence") are a central pillar in the development of new medical image analysis techniques. He organized the MICCAI grand challenges on WMH segmentation and brain tissue segmentation (MRBrainS13 and MRBrainS18). He developed freely available software for the detection of the midsagittal plane and surface and lesion-symptom mapping.

Prof. dr. Marcel Breeuwer

Marcel Breeuwer is full professor at the Biomedical Engineering and the Electrical Engineering departments of the Eindhoven University of Technology, The Netherlands, and Principal Scientist at the Magnetic Resonance department of Philips Healthcare, Best, The Netherlands. His research focuses on medical image analysis.

Prof. Dr. Sebastian Kozerke

Research area Sebastian Kozerke’s research interests include Magnetic Resonance imaging methodology with focus on ultra-fast dynamic imaging of perfusion, function and cardiac mechanics. A number of contributions to the field of k-t undersampling and parallel imaging methods have permitted important advances in spatiotemporal resolution and scanning speed for various applications. Other fields of activity concern probing of microstructure of moving organs using diffusion imaging methods. Recent research work also includes real-time imaging of substrate metabolism using dynamic nuclear polarization techniques. Curriculum Vitae Sebastian studied Electrical and Biomedical Engineering and obtained his PhD degree and the Venia legendi from ETH Zurich in 2000 and 2005, respectively. In 2002-03 he was a research associate at the Division of Imaging Sciences at King’s College London. In 2003 he co-founded the startup company GyroTools to help translating imaging research into applicable prototypes for Magnetic Resonance (MR) experimentalists and clinicians. He was elected as a Professor and Chair of MR Physics at King’s College London in 2008 before being promoted to Professor at the University of Zurich in 2010. In October 2014 he joined ETH as a Full Professor holding a dual appointment also at the Medical and Natural Sciences faculty of the University of Zurich. He is the chairman of EXCITE Zurich, a joint competence center to promote the development and application of experimental and clinical imaging technologies.

Sina Amirrajab

Sina Amirrajab is a PhD candidate in the Medical Image Analysis group of the Biomedical Engineering Department of the Eindhoven University of Technology. His main research interest is at the intersection of MRI and Artificial Intelligence, aiming to develop deep-learning solutions to tackle medical data scarcity. Since 2018 he has been involved in various projects related to cardiac MRI simulation and synthesis for generating training data for deep learning applications. Photo: The openGTN team at the ISMRM 2022 conference From left to right: Marcel Breeuwer, Aymen Ayaz, Yasmina Al Khalil, Sina Amirrajab

Aymen Ayaz

Aymen Ayaz is a PhD candidate in the Medical Image Analysis group of the Biomedical Engineering Department of the Eindhoven University of Technology. Her research interest is in the Neuro MRI domain and its applications for better diagnosis and prognosis. During her PhD, she has worked on projects related to Brain & Spine MRI simulation and synthesis to generate ample synthetic data, required for various deep learning applications. Photo: The openGTN team at the ISMRM 2022 conference From left to right: Marcel Breeuwer, Aymen Ayaz, Yasmina Al Khalil, Sina Amirrajab

Yasmina Al Khalil

Yasmina Al Khalil is a PhD candidate in the Medical Image Analysis group of the Biomedical Engineering Department of the Eindhoven University of Technology. Her research interests lie in the domain of computer vision and deep learning applications for medical image analysis tasks, with a specific focus on achieving more robust and adaptable MRI segmentation. During her PhD, she has worked on a number of projects related to utilization of synthetic data, as well as improved augmentation and adaptation strategies, in addressing various segmentation-related challenges in MRI. Photo: The openGTN team at the ISMRM 2022 conference From left to right: Marcel Breeuwer, Aymen Ayaz, Yasmina Al Khalil, Sina Amirrajab

Dr. Jakub Tomczak

Jakub M. Tomczak is an assistant professor of Artificial Intelligence at Vrije Universiteit Amsterdam and a founder of Amsterdam AI Solutions. Before joining Vrije Universiteit Amsterdam, he was a deep learning researcher (Engineer, Staff) in Qualcomm AI Research in Amsterdam, a Marie Sklodowska-Curie individual fellow in Prof. Max Welling's group at the University of Amsterdam, and an assistant professor and postdoc at the Wroclaw University of Technology. His main research interests include deep generative modeling, deep learning, and Bayesian inference, with applications to image processing, robotics, and life sciences. He is the author of the book entitled "Deep Generative Modeling". He has broad experience in consulting companies in logistics, banking, computer vision, and life sciences. He has over 12 years of experience working in academia and over 3 years working in and for the industry.

Dr. Cor Datema

Cor Datema is founder of the start-up company Dynaxion. Dynaxion is developing a new type of security scanner based on a particle accelerator that generates a beam of neutrons. After receiving his PhD in Physics from the university of Southampton, Cor worked in the early 2000s on the detection of landmines with neutrons at Delft University of Technology. After this he joined Philips where he worked in different roles in development (x-ray system design), marketing, supply chain management and sales & consultancy. In 2013 he founded his first start-up Recornect that provides psychiatric hospitals with an interaction screen for clients, loaded with various software apps for support during crisis, fully controllable by staff. Since 2019 he is the CEO of Dynaxion and plays an important role in the system design and development of the new scanning system.

Date

2 November 2022

Expired!

Time

10:30 - 18:30

More Info

Location

UMC Utrecht, collegezaal Blauw

Heidelberglaan 100, 3584 CX, Utrecht

Website

https://www.umcutrecht.nl/en/contact-directions

Speakers

Prof. Dr. Sebastian Kozerke

Prof. Dr. Sebastian Kozerke

Research area
Sebastian Kozerke’s research interests include Magnetic Resonance imaging methodology with focus on ultra-fast dynamic imaging of perfusion, function and cardiac mechanics. A number of contributions to the field of k-t undersampling and parallel imaging methods have permitted important advances in spatiotemporal resolution and scanning speed for various applications. Other fields of activity concern probing of microstructure of moving organs using diffusion imaging methods. Recent research work also includes real-time imaging of substrate metabolism using dynamic nuclear polarization techniques.
Curriculum Vitae
Sebastian studied Electrical and Biomedical Engineering and obtained his PhD degree and the Venia legendi from ETH Zurich in 2000 and 2005, respectively. In 2002-03 he was a research associate at the Division of Imaging Sciences at King’s College London. In 2003 he co-founded the startup company GyroTools to help translating imaging research into applicable prototypes for Magnetic Resonance (MR) experimentalists and clinicians. He was elected as a Professor and Chair of MR Physics at King’s College London in 2008 before being promoted to Professor at the University of Zurich in 2010. In October 2014 he joined ETH as a Full Professor holding a dual appointment also at the Medical and Natural Sciences faculty of the University of Zurich. He is the chairman of EXCITE Zurich, a joint competence center to promote the development and application of experimental and clinical imaging technologies.
Sina Amirrajab

Sina Amirrajab

Sina Amirrajab is a PhD candidate in the Medical Image Analysis group of the Biomedical Engineering Department of the Eindhoven University of Technology. His main research interest is at the intersection of MRI and Artificial Intelligence, aiming to develop deep-learning solutions to tackle medical data scarcity. Since 2018 he has been involved in various projects related to cardiac MRI simulation and synthesis for generating training data for deep learning applications.
Photo: The openGTN team at the ISMRM 2022 conference
From left to right: Marcel Breeuwer, Aymen Ayaz, Yasmina Al Khalil, Sina Amirrajab
Aymen Ayaz

Aymen Ayaz

Aymen Ayaz is a PhD candidate in the Medical Image Analysis group of the Biomedical Engineering Department of the Eindhoven University of Technology. Her research interest is in the Neuro MRI domain and its applications for better diagnosis and prognosis. During her PhD, she has worked on projects related to Brain & Spine MRI simulation and synthesis to generate ample synthetic data, required for various deep learning applications.
Photo: The openGTN team at the ISMRM 2022 conference
From left to right: Marcel Breeuwer, Aymen Ayaz, Yasmina Al Khalil, Sina Amirrajab
Yasmina Al Khalil

Yasmina Al Khalil

Yasmina Al Khalil is a PhD candidate in the Medical Image Analysis group of the Biomedical Engineering Department of the Eindhoven University of Technology. Her research interests lie in the domain of computer vision and deep learning applications for medical image analysis tasks, with a specific focus on achieving more robust and adaptable MRI segmentation. During her PhD, she has worked on a number of projects related to utilization of synthetic data, as well as improved augmentation and adaptation strategies, in addressing various segmentation-related challenges in MRI.
Photo: The openGTN team at the ISMRM 2022 conference
From left to right: Marcel Breeuwer, Aymen Ayaz, Yasmina Al Khalil, Sina Amirrajab
Dr. Jakub Tomczak

Dr. Jakub Tomczak

Jakub M. Tomczak is an assistant professor of Artificial Intelligence at Vrije Universiteit Amsterdam and a founder of Amsterdam AI Solutions. Before joining Vrije Universiteit Amsterdam, he was a deep learning researcher (Engineer, Staff) in Qualcomm AI Research in Amsterdam, a Marie Sklodowska-Curie individual fellow in Prof. Max Welling’s group at the University of Amsterdam, and an assistant professor and postdoc at the Wroclaw University of Technology. His main research interests include deep generative modeling, deep learning, and Bayesian inference, with applications to image processing, robotics, and life sciences. He is the author of the book entitled “Deep Generative Modeling”. He has broad experience in consulting companies in logistics, banking, computer vision, and life sciences. He has over 12 years of experience working in academia and over 3 years working in and for the industry.
Dr. Cor Datema

Dr. Cor Datema

Cor Datema is founder of the start-up company Dynaxion. Dynaxion is developing a new type of security scanner based on a particle accelerator that generates a beam of neutrons. After receiving his PhD in Physics from the university of Southampton, Cor worked in the early 2000s on the detection of landmines with neutrons at Delft University of Technology. After this he joined Philips where he worked in different roles in development (x-ray system design), marketing, supply chain management and sales & consultancy.

In 2013 he founded his first start-up Recornect that provides psychiatric hospitals with an interaction screen for clients, loaded with various software apps for support during crisis, fully controllable by staff. Since 2019 he is the CEO of Dynaxion and plays an important role in the system design and development of the new scanning system.