Carrefour de l'actualité

Québec Science Discovery of the Year: two Polytechnique Montréal professors among the top10 finalists

January 6, 2022 - Source : NEWS

No less than two Polytechnique Montréal professors have earned a top 10 spot in Québec Science magazine’s prestigious Discovery of the Year competition for 2021.

Professors Julien Cohen-Adad and Thomas Gervais.
Professors Julien Cohen-Adad and Thomas Gervais.

Associate Professor Julien Cohen-Adad (Department of Electrical Engineering), and Full Professor Thomas Gervais (Department of Engineering Physics), were both selected on the strength of articles considered to be exceptional by a panel of scientists and science journalists. The two professors were thrilled by the news.

“It’s obviously very gratifying overall, but what pleases me more is that the students who worked on the topic can also see the real impact of their research on society. It’s very rewarding, motivating and inspiring for students and research staff alike,” Professor Cohen-Adad comments.

Professor Gervais - who once contributed to Québec Science magazine as a freelance journalist - is flattered by this recognition, which he describes as “very special.” “I’m 100% committed to Québec Science’s mission of popularizing science carried out in Québec and around the world. The fact that our team’s work is a part of this worthy mission is a tremendous source of pride for us all.”

The Polytechnique Montréal community and general public invited to vote for the Professors and their work at .

Major step forward in spinal cord imaging

Thanks to the protocol developed by Professor Julien Cohen-Adad and his collaborators, AI can now be used on spinal cord MRI data. (Photo: NeuroPoly)
Thanks to the protocol developed by Professor Julien Cohen-Adad and his collaborators, AI can now be used on spinal cord MRI data. (Photo: NeuroPoly)

Professor Cohen-Adad’s work was selected after the publication of an article in Nature Protocols that could transform the world of magnetic resonance imaging (MRI) of the spinal cord, as explained earlier this year in this Labo 2500 article. MRI is a non-invasive imaging technique that has provided modern medicine with giant leaps forward by making it possible to see what’s happening inside the human body without surgery. It is used to detect spinal cord pathologies such as tumours or lesions associated with multiple sclerosis.

Yet until recently, one problem remained: spinal cord imaging data could be used only qualitatively. In the absence of reference points in the images, it was impossible to calibrate them to follow, for example, the evolution of a disease such as neuronal degeneration, or to measure the improvement of the condition of a patient undergoing treatment. The problem was such that images captured by the same device and/or technician a few months apart could not be compared.

Thankfully, that’s now a thing of the past thanks to the hard work and ingenuity of Professor Cohen-Adad and his team, who published their results in two high-impact articles in the Nature group series of magazines, including Nature Protocols.

The article details the steps of a protocol that make it possible to “standardize” spinal cord images taken by MRI, regardless of the device’s manufacturer. The group has tested its approach with 260 participants in 42 imaging centres around the world over the past five years.

"The standardized MRI acquisition method we propose can be used with almost all clinical MRIs in the world. By following this protocol, hospitals can ensure that all the data can be compared with other data, which will allow us to use an approach called ‘quantitative MRI,’ where we the size of microstructures is measured with high precision, to track their evolution over time,” Professor Cohen-Adad explains.

In addition to facilitating patient monitoring, this approach will enable specialists around the world to establish threshold values for detecting each spinal cord pathology. The team led by Professor Cohen-Adad, who notably is also a member of “Mila” an artificial-intelligence (AI) research institute, and they too will be lending a AI-hand with AI tools to improve diagnostic accuracy based on the imaging data acquired.

“A major limitation of artificial intelligence is the variability of data from different hospitals, which limits the generalization capabilities of deep-learning algorithms,” Professor Cohen-Adad notes. “With our discovery, images will be more standardized and therefore work better with new automatic analysis methods, such as for tumour detection in MRI images, for example.”

A liquid-pixel projector

Pixelated chemical display
The "pixelated chemical display" generates each colored pixel on its surface through a series of openings (gray dots) that are connected to a network of tubing where liquids are injected or aspirated. Rather than mixing, the liquids repel each other and create pixels. This is one of the amazing properties of liquid flows at the microscopic scale.

Professor Thomas Gervais and his team unveiled their “pixelated chemical display” - a kind of miniature liquid projector capable of conducting up to 144 experiments simultaneously. All this, on an area barely one square centimetre in size!

The aforementioned invention was reported in an article published in PNAS, and could transform the way we perform biochemical and biological tests. It may even bring a new type of robot into laboratories that use this so-called “microfluidic” approach, which relies on the physical properties of small volumes of liquids.

"To accomplish this innovation, we produced design rules for an entirely new kind of device that we’re pioneering. We believe that the publication of these design rules, in addition to the experimental demonstration, will enable many new microfluidic applications to emerge." says Professor Gervais.

The secret of the display lies in its ability to generate liquid “pixels” on its surface through a series of openings that inject or aspirate fluids. The patent-pending tool takes advantage of the amazing property of liquid flow on a microscopic scale: instead of mixing, the liquids repel each other to form each of the pixels.

The engineering team uses the display like a dynamic print. When in proximity to a surface, it transfers the fluids flowing across its surface point by point.

In a precursory paper published in Nature Communications in 2019, the same authors demonstrated how they performed the successive steps of an immunoassay using this approach, taking turns controlling the addition of reagents while conducting various wash cycles. In their most recent paper in PNAS, they demonstrate for the first time how their strategy could be replicated at will to conduct multiple experiments in parallel. The group demonstrated this by transferring proteins to a treated surface. They also proved the usefulness of their approach to fabricate flexible electronic circuits by depositing silver layers on plastic films.

“There’s still work to be done, but pixelated chemical displays could one day become a genuine laboratory replacement option for ‘pipetting robots,’ which are complex and expensive. These displays could eventually become a central component of liquid-manipulating robots that currently facilitate drug discovery, or they could even permit the development of new biomedical research approaches.”

Learn more

Professor Julien Cohen-Adad expertise
Professor Thomas Gervais expertise
Nature Protocols article by Professor Cohen-Adad
Article by Professor Gervais’s team published in PNAS
Department of Electrical Engineering website (In French)
Department of Engineering Physics website

Suggested Reading

May 14, 2014

Professor Julien Cohen-Adad edits a book on quantitative magnetic resonance imaging (MRI) of the spinal cord (Publisher: Elsevier)

June 22, 2021

Canada Research Chairs Program: Chair renewal for Polytechnique Montréal's Professor Julien Cohen-Adad