Projets novateurs réalisés

Photogrammétrie en conditions difficiles améliorée par des modèles d’intelligence artificielle

La photogrammétrie est une technique de modélisation où on utilise plusieurs images prises tout autour d’un sujet afin de le reconstituer en 3D. Parfois, les conditions de capture influencent la qualité de la reconstruction 3D : des obstacles obstruent certains points de vue, l’éclairage est mauvais, etc. De plus, quand le sujet est difficile d’accès, il n’est pas toujours possible ou économique de retourner prendre de meilleures images. Dans ces cas, les algorithmes qui analysent les images n’arrivent pas à calculer les paramètres des caméras pour les différents points de vue, nécessaires pour les assembler en 3D, et la reconstruction échoue ou comporte des défauts. Le CDRIN vise donc à développer, à l’aide de l’intelligence artificielle, un modèle capable d’améliorer le calcul des paramètres des caméras en photogrammétrie, plus robustes par rapport aux conditions de capture, et plus accessible au public parce que demandant moins d’ajustements et de configurations.

Conception d’une bibliothèque d’outils logiciels permettant la calibration et l’amélioration de la radiométrie des caméras infrarouge

Ce projet consiste en la conception, réalisation et tests d’outils informatiques servant a facilité le travail de recherche et développement de nouvelles méthodes de correction d’images digitales thermiques. Les outils crée par l’étudiant servirons à supporter les opérations d’acquisition de donnée et de calibrations et correction d’images infrarouges dans le but de les rendre visuellement propres (sans défauts) et d’être en mesure d’estimer la température en dégrée Celsius des objets dans la scène. Les mêmes outils serviront également à l’évaluation des performances de différentes méthodes ainsi qu’a la validation et au déploiement en production de masses. Ses outils permettront d’accélérer le développement et de diminuer les risques relier a la mise en production de caméras infrarouges (dites « thermique »).

Advanced Adaptivity and Personalization in Learning Systems through Collaborative Recommendations Year Two

Learning Systems are among the most popular e-learning tools in today’s education and training. Most e-Learning systems do not take into account individual aspects of learners (e.g., their goal, experiences, existing knowledge, learning style etc.).The primary goal of the proposed research is to offer rich adaptivity by combining information from a learner’s profile (e.g. levels, goals, learning style, cognitive abilities etc) with the information from other learners sharing common interests. Based on this combined information, advanced personalized recommendations can be provided, increasing efficiency, performance and learner’s satisfaction. The proposed research will have numerous benefits to the company: (1) Training and learning would become more accessible, to the benefit of employees in small to large -scale enterprises as it will offer unique learning experiences that fully engage and support users (2) It will help in improving and increasing the basic skills of employees, providing the organization with a competitive advantage and hence, will be used to build workforce capability

Retrospective analysis of operational legacy restoration

Alberta’s natural resource sector has resulted in the continuous and persistent loss of forest productivity. Despite huge strides in reclamation practices within the last 40 years, there is still a sizeable legacy of industrial disturbed sites that represent a significant opportunity to enhance the sustainability of the forest industry in the province.

Forest companies have been engaged in varying levels of reforestation of these disturbed sites but success has been mixed because often, the success of the program is not consistently assessed. This program presents a unique opportunity to evaluate forest regeneration success (or lack thereof) to address this knowledge gap between implementation treatments and the relative success of these treatments. The project’s core objective is to provide a quantitative understanding of forest recovery following an operational scale restoration program and to extract patterns in ecological region, ecosite or other attributes that may be used to predict forest regeneration ‘success’.

Developing machine learning tools for routine clinical EEG

The proposed project is part of a large-scale collaboration between SFU’s Behavioral and Cognitive Neuroscience Institute (BCNI) and Fraser Health Authority (FHA) in the domain of AI applied to clinical electroencephalographic (EEG) scans recorded and evaluated in the process of diagnostic workup in FHA’s public hospitals (n = 40’000). The key goal of the SFU/FHA collaboration is to automate the process of EEG reporting by building a decision support system. Conventional review of EEG relies on neurologists to visually inspect complex, noisy, high-dimensional digital data. Such an approach is slow, not fully reliable, and suboptimal. There is considerable variability in EEG interpretation, and this variability is affected by specific reader characteristics. Given these challenges, clinical reporting of EEG is highly suitable for machine-driven automation. The focus of the proposed project at a smaller scale is on predicting diagnoses (codes for most responsible diagnosis, secondary diagnosis, etc, according to International Statistical Classification of Diseases ICD-10) from in-patient EEG scans. We aim to address the following question: what current deep learning approaches can be used to reach state-of-the-art performance for EEG classification based on diagnostic codes.

INDÉPENDAMMENT ENSEMBLE : Projet de recherche-action comme levier à la création d’une oeuvre artistique interactive sur l’ÉDI (Équité, Diversité, Inclusion) en entrepreneuriat.

L’entrepreneuriat reste de façon largement prédominante présenté comme une activité purement économique, que ce soit dans la littérature ou les médias. Cette conception donne aussi l’image d’un entrepreneuriat réservé à une frange limitée de la population, négligeant certaines catégories de celle-ci. Des études ont ainsi mis en évidence la sous-représentation des femmes, des communautés migrantes, des membres de minorités visibles ou d’autres personnes ne correspondant pas à l’image stéréotypée d’« un entrepreneur ». Pourtant, pour cette population, l’entrepreneuriat représente souvent bien plus qu’une activité économique, et peut prendre place dans une dynamique profonde d’émancipation et de transformation sociale.
C’est cet entrepreneuriat ‘comme changement social’ que nous souhaitons mettre de l’avant dans ce projet de recherche-action, mené en partenariat avec l’organisme JOBEL North America. JOBEL est une organisation spécialisée dans la création d’oeuvres d’art interactives visant à faire découvrir à un large public (citoyens, organismes, entreprises, acteurs institutionnels, monde de l’éducation) des résultats de recherches scientifiques sur différents sujets de société.

Analyse des similitudes et disparités observées entre les recommandations de remboursements émises par les agences d’évaluation des médicaments et des technologies de la santé au Canada (INESSS et ACMTS) en fonction de leur champ thérapeutique

L’objectif de mon projet de recherche sera de faire ressortir les similitudes et les disparités concernant les recommandations de remboursements entre les agences d’évaluation des technologies de la santé au Canada (l‘INESSS pour le Québec et l’ACMTS pour l’ensemble des autres provinces) en fonction des différents champs thérapeutiques. Ce projet de recherche sera très pertinent pour l’ensemble des personnes travaillant en pharmacoéconomie et accès au marché pour mieux comprendre les tendances des agences de technologies de la santé, et ainsi mieux aborder leurs approches lors de l’élaboration d’un dossier de soumission selon le type de médicament. Ce projet pourra servir de guide de référence pour mieux comprendre les facteurs qui influencent la prise de décision et pour permettre de remettre des dossiers de soumission de qualité supérieur.

Carbonate Alteration Footprints of Hydrothermal Ore Deposits Year Two

Carbonate hosted hydrothermal ore deposits commonly develop narrow mineral alteration (i.e. visible) haloes, complicating exploration targeting. In contrast, hydrothermal modification of the country rock’s stable isotope composition usually extends far beyond visible alteration. Hence, stable isotope “mapping” is an effective tool to aid exploration for carbonate?hosted deposits. However, widespread utilization of stable isotope data has been hampered by its high cost and long turn-around times. Researchers at the Mineral Deposit Research Unit (MDRU) have recently developed an analytical instrument (the Mineral Isotope Analyzer, MIA) allowing for rapid and cost-efficient oxygen and carbon isotope analysis. The proposed research will assess how large-scale application of stable isotope analysis can be used to improve targeting in different exploration environments. The acquired data will help to constrain pathways responsible for controlling hydrothermal fluids, thus significantly further the understanding of how carbonate hosted hydrothermal ore deposits form, and can be more effectively targeted during exploration.

Screening DNA Methylation Signatures in Early Bovine In Vitro Embryos for Developmental Fitness

The success of assisted reproductive techniques (ART) in overcoming infertility or improving livestock genetic traits is undeniable. However, the use of ART in livestock has also been associated with an increased likelihood of certain anomalies (e.g., large offspring syndrome, musculoskeletal anomalies) and neonatal mortality because of suboptimal culture conditions. Our partner, L’Alliance Boviteq (Semex) has developed a series of tests (e.g., embryo morphology, genetic screens) to assess the quality of in vitro derived embryos. Still, a fraction of calves (less than 2%) is born with perinatal anomalies. In recent years, various studies have put epigenetic errors forward as a potential culprit for adverse offspring development following ART procedures. Epigenetic marks (e.g., chemical tags on the genome) are highly remodeled across the genome during the pre-implantation period and are key for regulating gene expression; including genes that have been associated with large offspring syndrome. Here, we aim to develop and implement a molecular screening approach based on epigenetic signatures to further improve prognostic ability of individual bovine embryos for developmental health prior to transfer.

Targeting Radiation-Induced CXCL12/CXCR4 Signaling and the Immune Microenvironment to Improve the Radio-Curability of Human Cancer

Medical images were collected from cervical tumors from mice undergoing cancer treatment using radiation to kill cancer cells and shrink tumors. The purpose is to discover new biological molecules and observe the immune system reacts during radiation treatment. These images will be analyzed by MATLAB, which is a computer platform. MATLAB will perform a couple of operations on an image to extract useful information from the image, such as the location and number of biological molecules and immune cells that appear during radiation therapy. By conducting the project, the project organization will gain a deeper understanding of new biological molecules that occur during radiotherapy and enhance the cure rate of radiotherapy.

Fermionic Molecular Dynamics simulations of quantum systems

In the last few decades, molecular dynamics computer simulations have become increasingly important to unravel and understand the mechanisms of (bio)molecular processes, significant for a myriad of pharmaceutical and/or industrial applications. Although these molecular systems should be considered on the microscopic quantum mechanical level, they are typically approximated by the macroscopic Newtonian laws because of fundamental computational limitations. Fortunately, whenever the quantum system is sufficiently large and close to equilibrium, the laws of thermodynamics become also applicable as they rely on statistical averaging procedures. In the present Mitacs Globalink project, we aim to investigate commonly accepted thermodynamical principles at a rigorous quantum level. We will develop and employ electronic Fermionic Molecular Dynamics (eFMD) simulations on a model electron gas to analyse quantum effects on the 2nd law of thermodynamics, stating that entropy (a measure of disorder in the system) should always increase over time.

Impact of environmental factors on spontaneous reporting trends and safety signal detection in the EudraVigilance database

Changes brought to the ecosystem and geopolitical environment can have an effect on several fields including the healthcare system. Recently, we observed the impact that COVID-19 pandemic has had on the healthcare workers’ practices and people’s behaviour regarding their health status. Changes in spontaneous reporting trends were also observed. Lately, Europe has also witnessed the Brexit which is another important event that can influence spontaneous reporting practices knowing that the United Kingdom is an important source of spontaneous reports. In Europe, adverse drug reactions (ADRs) suspected to be caused by authorized medicine can be reported to a publicly available pharmacovigilance system called EudraVigilance. This project aims at determining the impact of different environmental factors such as geopolitical changes (Brexit) and sanitary situations (COVID-19) on spontaneous reporting trends and practices from the EudraVigilance database. A context without the occurrence of these events will be simulated to develop a use case and assess if changes in reporting could have caused a loss in safety signal detection.