Sites Inria

Version française

European Research Council 2015

Anne Schneider - 27/04/2016

Rachid Deriche wins ERC Advanced Grant 2015

Rachid Deriche, an exceptional grade research director at Inria Sophia Antipolis - Méditerranée, where he heads the ATHENA project team, has been awarded a European Research Council grant in the Advanced category for the CoBCoM project, at the crossroad between mathematics, information technology, and neuroimaging.

ERC Advanced Grants are awarded under the "Excellent Science" pillar of the European Union's Horizon 2020 research and innovation programme, and are attributed to senior researchers recognised as leaders in their field and whose proposal for ground-breaking research has the potential to significantly push back the frontiers of science.

To tackle an extremely exciting field full of scientific challenges: exploring and understanding the last continent of uncharted territory, the human brain.

What does this grant represent for you? 

This prestigious EU grant is an extraordinary mark of recognition for the work I have accomplished so far, and a huge vote of confidence and support for the success of the project proposed.
It is a grant that salutes my research efforts for almost thirty years now in mathematical image analysis, computer vision, and, more recently, computational neuroimaging.

Receiving this grant and conducting this project for our Institute is a joy and an honour for me. I would like to take this opportunity to salute and extend my gratitude to all those I have had the pleasure of working with, especially my past students, the members of my Athena team, my colleagues at Inria, my friends at other national and international institutions, and of course my family to whom I am deeply indebted for their undying support and encouragement.

Concretely, how will this grant help your research?

In terms of both its duration and its amount, this grant is truly exceptional. Receiving this type of funding is simply a dream for any researcher who wants to tackle risky, long-term projects without having to worry about the financial aspects, which are very energy- and time-consuming.

Practically speaking, this grant will allow me to recruit a number of doctoral and postdoctoral researchers and engineers to join my team and help in developing and implementing the solutions that will allow us to reach goals we have set. It will also allow me to invite eminent colleagues, either to take part in certain conferences to be held or to spend time at Inria so that we can advance together and compare our solutions.

What has drawn you to this particular field of research for all these years?

In 2000, after about twenty years of research in imaging and computer vision, my scientific interest turned to the neurosciences, particularly the field of neuroimaging.

There were several factors that pushed me towards this shift in focus: at Inria, our research themes are established for a maximum period of twelve years, with an evaluation every four years by international experts from both academia and the industrial world. After three highly successful evaluations with my colleagues at the Sophia Antipolis centre in 1988, especially after creating the RealViz start-up in 1998 and the successful transfer of our research in 3D vision and imaging to this company, it was clear that I, with my colleagues—particularly Olivier Faugeras, with whom I have had the great pleasure of working over all these years—had an excellent opportunity to shift our focus and tackle a new field, to take on something far more complex and extremely exciting and full of scientific challenges: exploring and understanding the last continent of uncharted territory, the human brain! In the face of the daunting tasks that awaited us, my colleagues Maureen Clerc and Théodore Papadopoulo and I quite naturally created the Inria Athena project team in 2010, and Demian Wassermann joined us in 2014.

Was this really a new beginning?

In addition to all these reasons, I knew that I was going to live a unique and very exciting scientific experience, that of delving into a field in which the data you work with is heterogeneous, multi-scale, multi-modal, and highly complex, living on varieties such as the cortex; the experience of taking part in a fast-growing, high-potential field in the medical domain, of working your way up again in a new community where no one knows you, where the conferences and journals in which you need to be published are new, and where you have to build a new network of academic, clinical, and industrial partners in order to create the algorithms behind concrete applications in the neurosciences and neuroimaging.

In short, to step back into the shoes of the young researcher I had been when I started out in mathematical image analysis and processing, and to stretch my cerebral faculties again as I had when I starting working on computer vision. All of this seemed to correspond exactly to what I was looking for at the time. All said and done, I certainly have no regrets about that decision.

What is the aim of the research proposal that you submitted to ERC?

The aim of the CoBCoM research project, short for "Computational Brain Connectivity Mapping", is to construct a dynamic network of the structural and functional connectivity of the human brain.

To construct a dynamic network of the structural and functional connectivity of the human brain and pave the way to developing new markers.

This will involve developing new imaging and analysis techniques based primarily on the data produced by Diffusion-Weighted MRI for the structural aspect, and by Magnetoencephalography (MEG) and Electroencephalography (EEG) for the functional aspect, and then applying them in order to dynamically map the brain connectivity network and pave the way to developing new cerebral imaging markers.

What is unique about your approach?

The imaging techniques that we work on (DW-MRI, MEG, and EEG) enable non-invasive, in-vivo exploration of the brain with good spatial resolution (DW-MRI) and good temporal resolution (MEG and EEG). It can be said that Diffusion-Weighted MRI is the only technique that allows us to travel along the white matter fibre bundles and non-invasively retrace the information highways of the human brain, and that EEG and MEG are the premier techniques for non-invasive exploration of brain activity.

What is the scientific challenge?

There are a number of problems with acquiring data through DW-MRI, EEG and MEG, mainly in terms of the quantity, complexity, and heterogeneity of the data. Building on the modelling and methodological developments based on mathematical and IT tools that offer both variational methods for mathematical signal analysis and processing and advanced signal learning and processing techniques, what we are going to do with this project is to take modelling even further by developing a dynamic network of structural and functional brain connectivity.

To accomplish these feats, we will use higher-order models and Riemannian geometry and graph theory tools, also tying in the microstructure data associated with the white matter fibre network and the physics of signal formation. At its core, this scientific project is methodological, but of course my team and I will also interact and collaborate with a variety of disciplinary fields. With the help of our partners in the clinical field, we will also strive to ensure that our work is concretised in real practical benefits, leading to the development of new markers in cerebral imaging and neurodegenerative diseases.

In conclusion

This research has significant societal and economic impact. A recent European study estimates that more than a third of the European population is affected by what we call mental disorders.

While health has no price, it does have a cost, and with the increasingly ageing population, there are clearly enormous interests at stake in developing techniques for better understanding and diagnosing neurodegenerative diseases.

Keywords: Rachid Deriche Bourse ERC Inria - Sophia Antipolis - Méditerranée EPI Athena Award CoBCoM H2020 Neurosciences EEG IRM de diffusion MEG

Top