Two ERC Advanced Grants

The European Research Council has bestowed Advanced Grants on two members of the School of Engineering, in addition to the two Starting Grants announced in November of last year. Advanced Grants are awarded to projects which the ERC consider to be "highly ambitious, pioneering and unconventional". The recipients are Prof. Michael Unser, of the Biomedical Imaging Group, and Prof. Nava Setter of the Ceramics Laboratory.

In 2010 the Dean of the School awarded a research prize (in the form the funding of a Ph.D. student over 3 years) to the two laboratories who were judged to be the most productive, based on the number of publications per full-time equivalent. The recipients were the same two labs as have been recognised by the ERC.

Prof. Setter’s Ceramics Laboratory brings together several highly recognized scientists in the field of ferroelectrics, embodying complementary competences from theory and modeling, through processing-science and structural and functional characterization, to design, fabrication, and testing of new types of micro- and nano-devices based on ferroelectrics.

In interview, Prof. Setter explains in detail how this research grant will affect the research of the Ceramics Laboratory:

    Ferroelectric materials contain interfaces called domain walls. These interfaces, which have the typical thickness of 1-2 nanometers are mobile. In principle they can be controlled and modified dynamically by external forces (electrical, stress fields, temperature variations) and even can be annihilated and recreated. Our dream is to functionalize them, making devices that can be modified and repositioned dynamically ("Device on-demand"). Such devices could potentially introduce new paradigms in electronics and in sensing.

    Because of their small thickness and instability, ferroelectric domain- walls are only vaguely understood and hardly controlled at present. Instrumentation that allows one to probe individual domain walls has been developed only in last decade. Last month, the Swiss National Science Foundation has granted our laboratory the funds to acquire a most advanced tool for probing of domain walls; it will be the second such equipment available world-wide.

    The ERC grant (and additional support) will allow our team of the Ceramics laboratory to focus on ferroelectric domain walls: to investigate their fundamentals and then hopefully to be able to stimulate and tame them, leading potentially to new functionalities. Such an extensive effort would be impossible without the ERC grant.

How do you think the fields of ferroelectrics and piezoelectrics will change over the next 50 years?

    I think the key issue is to be able to ‘control’ the material. This is the essence of the profession of the Materials Engineer altogether:

    Will we be able to ‘engineer’ new materials, which perform better than the ones available at present? There are many reasons to believe this is possible. The benefits will be many; For example, stronger piezoelectric materials will enhance the resolution of ultrasonic medical imaging, making it an extremely powerful technique for health monitoring.

    Another dream is to merge magnetism, ferroelectricity and semiconduction. At present semiconductors are controlled minutely in both their composition and properties. Ferroelectrics are more complex, as typically their properties are tied with instabilities. Ferromagnetic materials have some similarity to ferroelectrics, but in a way they are simpler and better understood. Merging of the three might lead to a new type of high-efficiency electronics (further miniaturization, higher operation frequency, etc.). There is a potential for an impact on various fields: energy, medical diagnostic instrumentations, remote communications, and so on, but all this is at present only a vague dream.

Prof. Unser’s group works at the theoretical and practical frontiers of biomedical imaging and signal processing:

    Firstly, we work on fundamental issues such as the representation and modeling of images which depend on rather sophisticated mathematics: functional analysis, approximation theory, and splines. With the current ERC grant (FUN-SP: A functional framework for sparse, non-Gaussian signal processing and bio-imaging), we aim at contributing to the statistical foundation of signal processing by formulating the "sparse" counterpart of the classical theory of Gaussian stochastic processes.  We believe that such a theoretical framework is required to support the current dominant trend in signal processing and imaging: the development of second-generation algorithms based on the concept of sparsity. Building on our expert knowledge of splines, we plan to use partial differential equations and distribution theory as the corner stone of our approach.

    Secondly, in parallel to the theory, we work on real practical imaging problems in close collaborations with biomedical scientists. On the one hand, we are developing novel reconstruction algorithms for specific modalities (MRI, fluorescence microscopy, digital holography, phase-contrast tomography) in direct collaboration with the imaging scientists who are designing the instrumentation. On the other hand, we are developing public domain image analysis software (as plugins for ImageJ) to respond to the direct needs of the users (biologists).  We are aiming at having a real practical impact, while maintaining a high level of signal processing. 

    Winning the ERC grant will give us the means to pursue  the above objectives. We will use the fund and prestige of the ERC grant to attract the exact type of collaborators we need: applied mathematicians who are willing to collaborate with engineers, and signal processing researchers with hands-on expertise in biomedical image analysis.

How do you see the future of Biomedical Imaging?

    The computational part of imaging (signal processing, sophisticated reconstruction algorithms) will take a more dominant role, while the imaging instrumentation will get streamlined and scaled-down (miniaturized). The current limits of resolution will be overcome through the development of new paradigms and novel imaging probes. Functional (as opposed to structural or anatomical) imaging will become more prominent. The handling of larger amounts of data will require intelligent forms of processing and better archival strategies.

    Medical imaging: the push will be towards micro-imaging using enhanced version of MRI and  coherent X-ray tomography (including phase contrast). Radiology will be primarily based on differential analysis; that is, the computer-assisted detection of changes with respect to some reference atlas or/and some earlier examinations. 

    Biological imaging: Optical and fluorescence microscopy, which is currently under development, will reach its golden age. The physical Rayleigh limit will be overcome by using non-linear excitation schemes together with improved fluorescent probes, which will allow researchers to explore the uncharted range of (molecular) resolutions between 200-20 nm.

    Biologists will finally have access to 3D time and multispectral data sets over several orders magnitude and will be able to visualize them using image navigators akin to "google earth".

    Imaging will remain as important as ever and will contribute significantly to the advancement of science. It is likely (once more) to qualify as one of the key technologies of the 21st century.

Nava Setter completed her MSc in Civil Engineering in the Technion (Israel) and her PhD in Solid State Science in Penn. State University (USA). She began her affiliation with EPFL in 1989 as the Director of the Ceramics Laboratory, becoming Full Professor in 1992. She has been Head of the Materials Department in the past and is currently Director of the Doctoral School for Materials. She is a Fellow of the Swiss Academy of Technical Sciences, the IEEE, and the World Academy of Ceramics.

Michael Unser received the M.S. (summa cum laude) and PhD degrees in Electrical Engineering in 1981 and 1984, respectively, from the EPFL. He spent 13 years working as a scientist with the National Institutes of Health, Bethesda MD, before being appointed professor at the EPFL in 1997. His main research topics are biomedical image processing, splines and wavelets. He has published 200 journal papers, and is one of ISI’s Highly Cited authors in Engineering. Prof. Unser is a fellow of the IEEE, and a recipient of three IEEE-SPS Best Paper Awards.