Consistent with the mission of the CIBM to teach and educate its membership, in 2021 a grassroots effort in the form of a grant writing workshop organized by Dr. Cristina Cudalbu and Dr. Meritxell Bach Cuadra is now bearing fruit.
Not one but three CIBM core members have been awarded an SNSF grant following the project funding calls MINT and Life Sciences in October 2022.
It is noteworthy to mention that this is a significant achievement for the two CIBM research staff scientists, Dr. Tomas Ros and Dr. Jérôme Yerly, as it is their very first SNSF grant award which will allow them to establish themselves as independent researchers and lead their own team for the next 4 years.
The CIBM Leadership congratulates the three awardees and acknowledges their contributions in achieving CIBM’s strategic objectives.
SNSF MINT 2022 Awardee: Meritxell Bach Cuadra
Title: Tackling domain shifts in paediatric neuroimaging: bridging advanced computational MR techniques and clinical practice
Partners and Collaborators:
- CHUV: Vincent Dunet, Patric Hagmann, Mériam Koob, Juliane Schneider, Anita Truttmann
- Siemens Healthineers: Tobias Kober
- King’s College London: Maria Deprez, Jana Hutter
- CIBM MRI CHUV-UNIL: Matthias Stuber
- CIBM SP EPFL: Pol Del Aguila Pla, Michael Unser.
Amount: 969,531 CHF
Keywords: Information Technology, Electrical Engineering, Neurophysiology and Brain Research, Paediatrics, Biomedical Engineering
A few thoughts from Meritxell
This is our 4th SNSF funded project (since 2012) related to advanced image processing techniques for fetal neuroimage analysis. It ensures the continuity of our developments/research and allows us to face new challenges and exploit new opportunities raised by low field MRI.
The aim of our research is to reduce two major domain gaps in pediatric MRI, linked to different magnetic field strengths and populations. To this end, novel MR imaging super-resolution reconstruction and segmentation methods robust to distributional shifts will be developed.
Today, machine learning methods applied to brain MRI at 1.5T or 3T enable the extraction of imaging biomarkers of the early brain maturation and their association to later cognitive-psychological conditions and behavioral changes. Assessing the impact of brain development in fetal, preterm and newborn populations later in life is however limited by the domain shift between pre- and postnatal imaging techniques. Furthermore, at the global level, existing neurodevelopmental studies are also skewed by excluding a wide part of the population due to limited access to MRI, thus narrowing our understanding of brain development in normal conditions as well as in situations of malnutrition, infection or stress, which are predominant in middle- and low-income countries. The recent push for low-field MRI could be the game changer towards a global dissemination of pediatric neuroimaging.
As with the advance of machine learning and MR imaging technologies, it is crucial and timely to ensure their robustness to domain shifts. Our developments would provide the physicians and neuroscientists the possibility to quantitatively assess maturation processes via morphological and relaxometry biomarkers in a continuum between pre- and postnatal periods, independently from the used MR system, towards democratization of pediatric MRI.
The developed machine learning methods towards generability and improving low-field structural MRI can also be of interest for other MR imaging modalities and organs.
CIBM SP CHUV-UNIL section has recognized expertise in advanced medical image processing, especially in super-resolution reconstruction, segmentation and machine learning. The project partners have an extensive expertise in pediatric and fetal neuroradiology, MRI physics, sequence development, and quantitative imaging. Uniting these different collaboration partners in the context of this project therefore provides a perfect setting for the expected breakthroughs towards advanced machine learning methods robust to domain shifts for the assessment of fetal and neonatal brain development and pathologies.
SNSF Life Sciences 2022 Awardee: Tomas Ros
Title: Innovating neurofeedback therapy for ADHD: training real-time brain (micro)states
Partners and Collaborators:
HUG Nader Perroud, Marie-Pierre Deiber, Roland Hasler
UNIGE Victor Férat
University of Tübingen Thomas Wolfers
Utrecht University Martijn Arns
Amount: 761,489 CHF
Keywords: EEG, Brain-Computer Interface, neurofeedback, ADHD, Psychiatry
A few thoughts from Tomas
This is my first major grant and it will enable me to be Principal Investigator in my area of specialization, which is closed-loop brain training (i.e. EEG-based neurofeedback.
The project will develop an EEG platform to provide feedback in a closed-loop based on real-time estimates of their brain (micro)state activity. This may inspire future medical researchers to collaborate with the CIBM based on this pioneering approach.
The current front line treatment for attention-deficit hyperactivity disorder (ADHD) is pharmacotherapy, but this is associated with decreased efficacy over time as well as adverse side effects. Our translational project aims to clinically validate a non-pharmacological therapy called neurofeedback, which enables direct training of brain activity through a non-invasive brain-computer interface. This approach is novel as we will specifically target the neural correlates of the psychiatric disorder, based on the real-time dynamics of a patient’s brain (micro)states. The microstate framework takes into account both the rapid (i.e. sub-second) brain dynamics necessary to support cognition, as well as the topography of multi-channel EEG, which enables a more accurate discrimination of the diverse anatomical generators of brain activity.
The CIBM EEG HUG-UNIGE Section led by Prof. Christoph Michel has invested many years in exploring EEG microstates as potential biomarkers of psychiatric and neurological disorders.
My recent work has shown that patients with ADHD demonstrate abnormal microstate patterns that are replicable when compared to neurotypical controls. This project is the fruit of the fusion of the microstate framework with the concept of closed-loop brain training (i.e. neurofeedback).
The CIBM EEG HUG-UNIGE Section is open to collaborate with other parties (e.g. Synapsy Centre) in order to apply neurofeedback training in a wider range of psychiatric/neurological disorders. This could be done by applying for additional funding and/or organising scientific conferences on the topic.
SNSF Awardee: Jérôme Yerly
Title: Detailed Structural Imaging And Comprehensive Characterization Of The Arrhythmogenic Myocardial Substrate Using Super-Resolved Free-Running Multidimensional Cardiac Imaging (MERCATOR)
Partners and Collaborators:
- IHU LIRYC Aurelien Bustin, Hubert Cochet, Pierre Jais
- CHUV-UNIL Ruud van Heeswijk
Amount: 431,289 CHF
A few thoughts from Jérôme
This SNSF research project grant will provide me with the critical initial independence that is needed to achieve my long-term professional goal of establishing an independent research group that focuses on developing novel non-invasive imaging methodologies to better diagnose and guide therapy of cardiovascular diseases. It will allow me to independently collect new data, to open new avenues of research and discovery in the domain of cardivascular diseases, and to disseminate my results at international conferences and in high-profile journals as the senior author. It would furthermore enhance my supervision and teaching experience by giving me the opportunity to fully guide a PhD student.
For the team members of the CIBM MRI CHUV-UNIL Section, it will contribute to our overarching vision of improving accessibility of cardiovascular magnetic resonance for patient care.