Innovative Projects Realized

Découpe virtuelle interactive de corps élastiques déformables par une méthode de particules

La simulation chirurgicale offre un environnement de pratique sécuritaire pour la formation de nouveaux chirurgiens. De nombreux défis techniques doivent être relevés pour élaborer un environnement virtuel réaliste et utile. Ce projet aborde en particulier la simulation du comportement biomécanique des organes lors de leur interaction avec l’utilisateur, avec une attention particulière apportée au problème de la découpe.
Nous voulons simuler de façon précise les déformations et les forces ressenties par les organes lors de leur contact avec les outils chirurgicaux. Les approches les plus courantes reposent sur la méthode des éléments finis, qui représente les organes en les divisant en éléments géométriques interconnectés. Cette interconnection rend difficile la simulation d’une incision sur un organe étant donné qu’elle nécessite de créer de nouveaux éléments pour remplacer ceux qui se font découper. En contraste, nous voulons représenter les organes avec un nuage de particules non connectées, ce qui rend beaucoup plus facile la simulation d’une incision. Puisque les méthodes avec particules sont souvent plus lentes que celles avec éléments finis, nous utiliserons une carte graphique (GPU) pour accélérer les calculs.

Nouvelle approche pour l’optimisation quadratique sous contraintes de bornes

Le projet de recherche concerne l’étude et de développement d’algorithmes pour l’optimisation de fonctions différentiables sous de simples contraintes de bornes. Ces problèmes présentent un intérêt intrinsèque par leur structure. Des algorithmes d’application plus générale utilisent fréquemment des sous-problèmes avec contraintes de bornes. Plusieurs applications, par exemple dans les problèmes d’imagerie, comportent systématiquement des bornes de non-négativité correspondant à la réalité physique qu’un pixel à valeur négative n’a pas de sens.

Depuis la cinquantaine d’années que la discipline de l’optimisation mathématique a pris son essor, l’intérêt pour les contraintes de bornes a été maintenu au fil des progrès dans le domaine algorithmique. C’est l’esprit de ce projet de recherche, d’exploiter des résultats modernes sur les algorithmes pour l’optimisation sans contrainte dans le traitement des bornes. Une piste d’algorithme avec bornes sera développée, un algorithme sera implémenté et validé au sein du logiciel Qpalm.

Bounds on LDPC codes

Quantum circuit components will always be unreliable. To protect quantum information from becoming corrupted, we require quantum error correcting codes. The drawback is that quantum error correcting codes necessitate a trade-off – the better the code protects information, the more resources it requires to be sustained. Our current resource estimates to construct useful quantum circuits seem insurmountable.
It was recently shown that if a certain class of error correcting codes, called quantum LDPC codes, were to exist, then they could potentially lower resource requirements significantly. However, it is unknown whether these codes exist and the hunt for these codes has become an active area of research.
By borrowing ideas from the theory of classical error correcting codes, we hope to explore the limits of quantum LDPC codes. These methods are abstract yet powerful tools that exploit the mathematical structure of LDPC codes to bound their performance. By establishing what is (im)possible with quantum LDPC codes, we help move towards answering an important open problem towards the realization of quantum computation.

A unique and continuous tractography for both brain and spinal cord

As a part of work between the hospital of Rennes and the VisAGeS team, the EMISEP project focuses on multiple sclerosis (MS) which affects the spinal cord. We address the evaluation of the potential of focal lesions and diffuse lesions observed using diffusion MRI early in the disease course to predict the EDSS at 5 years in relapsing-remitting MS patients. The diffusion MRI data will be pre-processed prior to the beginning of the internship: distortion correction, diffusion model estimation, fiber tracking of the brain and spine separately. We will focus on the development of a strategy to join both tractograms to obtain a continuous representation of fibers for brain and spinal cord. We will first evaluate state-of-the-art registration method to identify the main challenges in the task of gluing together brain and spine tractography. The processing will be performed on the whole EMISEP dataset, and the results will be evaluated based on the total length of the reconstructed fiber bundles. Last, the results of fiber tracking will be used to perform track-based statistical analysis of the diffusion parameters along the reconstructed fibers and correlated with clinical scores.

Interoperability of proofs with Dedukti and Beluga

Critical systems such as transportation systems require a high level of safety that can only be achieved with formal proof. Such formal proofs are typically expressed in some logic that can be verified by theorem provers. The diversity of theorem provers and logics has a negative consequence: the same theorem is proved many times and it is difficult for these systems to co-operate, because they do not implement the same logic. Logical frameworks are a class of theorem provers that overcome this issue by providing a generic framework in which we can represent and specify various logics. The logical framework Dedukti developed at INRIA shines when it comes to compactly represent logics using user-defined rewrite rules, but lacks the ability to write proof transformations between logics. The logical framework Beluga developed at McGill excels in writing such proof transformations, but does not allow user-defined rewrite rules. This project aims at building a new logical framework combining the strengths of Dedukti and Beluga. Concretely, we plan to first design a unified logical framework that supports both user-defined rewrite rules and the ability to write proof transformations. TO BE CONT.

Deployment of wireless sensor networks for air pollution monitoring

Air pollution is one of the main environmental applications where Wireless Sensor Networks (WSN) are widely used. Using WSN for air pollution monitoring usually targets two main applications: 1) regular mapping and 2) the detection of high pollution concentrations. Both of these applications need a careful deployment of sensors in order to get better knowledge of air pollution. We have already proposed for both applications novel deployment approaches allowing to minimize the deployment cost of the monitoring system while ensuring an effective monitoring of air pollution and maintaining the network connectivity. Nevertheless, in our existent work, we do not take into account the energy consumption of nodes and the effective routing and collection of pollution data. Also, our solutions are not adapted to 3D regions. The aim of our collaboration project is therefore to cope with these limitations while integrating into our approaches the aspects of energy consumption, effective data collection and routing in addition to 3D deployment. Thanks to the remarkable knowledge of the host Professor A. Boukerche through his numerous contributions in the field of WSN deployment and data collection, we believe that our collaboration would be fruitful.

Modeling the Risks and Damages from a “Potential” Invasive Plant Species: Yellow Starthistle in British Columbia

The purpose of this research project is to forecast the timing and estimate the costs of yellow starthistle (Centaurea solstitialis) invasion into southern British Columbia. Yellow Starthistle is an invasive plant that has caused tens of millions of dollars of damages to agricultural production in the United States as well as millions of dollars of costs in the form of reduction of soil moisture, losses of biodiversity and tourism. YST has been detected in states immediately adjacent to the Canadian border (Washington & Idaho), and due to climate change, the spread of YST into Southern British Columbia and Alberta is not a question of if but of when and where. Based on an extensive literature review pertaining to invasive species prevention and control, and bio-economic modelling we will outline a set of policy recommendations that minimise the risk of invasion and therefore control costs. The above recommendations will be passed on to the invasive species societies whose mandate is public education, as well as prevention and control of biological invasions.

Favoriser la créativité et l’innovation par les espaces physiques

Dans un but de développement économique et de revitalisation du centre-ville de Sherbrooke vient s’inscrire le projet du quartier de l’entrepreneur : Well inc. Un quartier dédié au développement entrepreneurial à Sherbrooke. Les enjeux principaux sont de fournir à la communauté et aux entrepreneurs des services, des ressources et des outils pour les soutenir dans la création et le déploiement de leur entreprise. Le «Bâtiment A» a été identifié comme le lieu qui regrouperait différents espaces pour les entreprises et les organismes de développement économique et au sein duquel le contact de ceux-ci et de la communauté serait possible. Les responsables du projet imaginent actuellement l’aménagement physique de ce bâtiment. Ce projet de recherche souhaite étudier les environnements physiques qui favorisent la créativité et l’innovation ainsi que la collaboration et les interactions. Nous pourrons étudier un phénomène qui s’inscrit dans le caractère unique qu’est le développement Well inc.

Microfluidic Point-of-Care Immunoassay

In this proposed project, a collaboration between Tel-Array Diagnostics and the UBC research team
will result in the development of a fully automated and portable point-of-care test for breast cancer
proteins from a single drop of blood. In order to achieve this ambitious goal, we will develop, optimize
and validate a fully integrated microfluidic sample processing and imaging system, specific for Tel-Array immunoassays. The intern will optimize the incubation times needed for storage, preparation,
sample processing and experiment readout for the developed test. The intern will also validate the test
against standard laboratory tests to determine its specificity and sensitivity. The outcomes will also
provide a platform technology for Tel-Array to launch products into other areas of personalized
medicine, such as heart disease and diabetes.

EEG Signal Processing as a Predictor of Anti-Depressant Response

Major depressive disorder (MDD) is a serious mental condition that often completely debilitates a

client. MDD is typically treated with one of several currently available antidepressant medications.

However, the response rate to any of these medications is only about 30%. Unfortunately, there are

currently no means for a priori assessment of whether a specific person will respond to a particular

medication. Thus, in prescribing a treatment for MDD, the psychiatrist must by necessity resort to a

trial-and-error procedure. This can result in long delays before remission and significant stress on the

health care system.

In conjunction with collaborating psychiatrists, the applicant has developed a preliminary EEG-based

machine learning (ML) methodology that can predict the response of a person to an SSRI medication

(which is one of the classes of anti-depressant treatment) before the therapy begins. It is clear that

such a capability, when fully developed, will vastly improve the treatment of MDD. However, before

this system can be exploited …

A One-Health approach to echinococcosis (Echinococcus multilocularis) in client owned dogs in Alberta: force of infection, risk factors, diagnostics, treatments and information

This project aims to determine prevalence rates of Echinococcus species tapeworms in domestic dogs and to compare these rates to those obtained from wild canids including wolves, coyotes, and foxes. This will shed light on infection rates and risk factors for both the human population and their pet dogs. This research will benefit Bayer as they produce many pharmaceuticals that combat Echinococcus infection and believe an increased public awareness of this parasite is important for public health. Adequate education on Echinococcus will help people to avoid this risk and also to recognize potential infection in their homes.

Bio-refinery of wastewater treatment in presence of CO2 utilizing micro-algae

A micro-algae bioreactor or photobioreactor can be used for cultivating micro-algae for the purpose to
fix C02 or produce useable and commercial by-products. Specifically, micro-algae bioreactors can be
used to produce fuels such as biodiesel and bioethanol, as a replacement to conventional fuel sources.
The production of these biofuels also reduces greenhouse gasses as part of the growth cycle of microalgae.
Fundamentally, this kind of process is based on the photosynthetic reaction which is performed
by the chlorophyll-containing micro-algae itself using dissolved carbon dioxide and sunlight energy
though other literature has also studied micro-algae growth in the absence of light which is called
heterotrophic reaction .