Projets novateurs réalisés

Adaptive ML-Driven Detection of Scheduled Task Anomalies and Automated Threat Attribution

As cyber threats grow more sophisticated, attackers increasingly exploit scheduled tasks to maintain persistence and evade detection. Traditional security measures struggle to distinguish between legitimate and malicious task executions, especially when attackers modify execution parameters. Additionally, identifying and attributing threats to known adversaries remains a complex and resource-intensive process, relying heavily on human analysts and labeled data. This project, in collaboration with eSentire, aims to bridge these gaps by developing:
– An ML-driven anomaly detection system to identify suspicious scheduled task executions based on execution flow anomalies, registry modifications, and stealthy command-line manipulations.
– An automated threat actor attribution pipeline leveraging semi-supervised learning and the Diamond Model of Intrusion Analysis to enhance the accuracy, speed, and scalability of adversary identification.
By advancing cybersecurity analytics, this research will strengthen eSentire’s proactive threat detection capabilities, improve SOC efficiency, and contribute to industry-wide efforts in tackling automation-based persistence techniques and adversary attribution.

Dissecting the role of PCSK9 in sepsis

Severe sepsis strikes young and old alike with an increasing incidence of >75,000 per year in Canada at a cost of $40,000 per patient. In 40% sepsis is complicated by low blood pressure and organ failure with a mortality rate of 30-60%. The number of deaths due to severe sepsis and septic shock is greater than the number of deaths due to acute myocardial infarction. The process that links infection to organ failure and death is triggered by bacterial toxins, including lipopolysaccharide, lipoteichoic acid, and phospholipomannan. Pathogen toxins are cleared from the blood by the liver. We have recently discovered that inhibition of Proprotein Convertase Subtilisin/Kexin type-9 (PCSK9) improves survival in septic shock, possibly by altering clearance of pathogen toxins. This research will discover how best to neutralize PCSK9.

Sex differences and risk factors

It has been demonstrated that hypertension, type II diabetes, obesity, high cholesterol, and smoking are vascular risk factors that increase the chances of experiencing accelerated cognitive decline or developing dementia. Some studies have found that these vascular risk factors affect women’s cognition more severely than men’s. However, the reasons behind these observations remain to be elucidated. Some researchers have hypothesized that the decrease in estrogen levels during menopause increases women’s risk of developing vascular risk factors and experiencing accelerated cognitive decline. Hormonotherapy might compensate for this estrogen reduction. However, the complex interaction between sex, vascular risk factors, and cognitive decline, as well as the role of menopause and hormonotherapy in this interaction, still needs to be understood.

Our study aims to identify sex-related differences in the impact of vascular risk factors on cognitive decline and to better understand the role of menopause and hormonotherapy in this interaction by using data from the Canadian Longitudinal Study on Aging. This study will provide more knowledge about the impact of vascular risk factors in cognitive decline and the risk of developing dementia, but also contribute to a better understanding older women’s health, who face a double stigma in research, ageism and sexism.

Ultrafast Probing of Defect-Engineered 2D Quantum Materials

Two-dimensional materials (2DMs), such as graphene, have revolutionized the field of nanotechnology due to their exceptional electronic, optical, and mechanical properties. However, the introduction of controlled defects into these materials offers an exciting avenue for tailoring their functionalities for next-generation technologies. This project focuses on studying the effects of defects, such as grain boundaries and vacancies, on the structural and electronic behaviour of 2DMs.

The research involves synthesizing defect-engineered 2DMs using advanced epitaxial growth and chemical vapor deposition (CVD) techniques. State-of-the-art characterization methods, including high-resolution transmission electron microscopy (HRTEM), femtosecond electron diffraction (FED), and ultrafast transient absorption (TA) spectroscopy, will be used to investigate how these defects influence carrier dynamics, lattice vibrations, and optoelectronic performance.

By combining experimental techniques, this project aims to establish clear correlations between defect structures and material behavior, providing valuable insights into how defects can be harnessed for designing improved materials. The results will have broad applications in areas such as photodetectors, flexible electronics, and quantum devices. This research contributes to advancing 2D material science, paving the way for innovative technologies that rely on defect engineering.

Energy Audit Reporting Automation with Artificial Intelligence (AI) and Machine Learning (ML)

Currently, energy audits at Dillon Consulting Limited take 2 to 3 days to generate comprehensive reports, consuming significant auditor time due to the manual process. This inefficiency diverts auditors from critical tasks and introduces inconsistencies, especially in buildings with multiple facilities. It limits Dillon Consulting’s ability to meet increasing demand for energy management practices. There is a critical need for advanced AI techniques capable of processing inputted data, generating human-quality reports, and adapting to evolving needs. This project addresses this gap by developing an automated system using AI to transform the manual audit report process. The proposed system automates data transfer from site visits into pre-formatted Word templates using AI to generate unique descriptions. By leveraging past reports for references, the software will retrieve relevant data for similar building types and systems, automating 75–80% of the report. Auditors will only need minimal adjustments, reducing workload and improving consistency. Over time, the AI tool will adapt and improve with new data, creating a robust, efficient, and scalable solution for energy auditing.

AN ADAPTIVE PHASE I INTRA-PATIENT DOSE ESCALATION STUDY OF FENRETINIDE IN ADULT CYSTIC FIBROSIS PATIENTS

The proposed research project is a Phase 1 clinical trial looking at the safety of fenretinide, a molecule based on the structure of vitamin A, on patients with a disease called cystic fibrosis. Fenretinide has been shown in our laboratory to help cystic fibrosis mice by lowering molecules involved in inflammation and increasing the levels of docosahexaenoic acid (DHA), which is a fatty acid known to decrease inflammation. The intern will be involved in patient sample processing, analyzing and logging. She will also work with Laurent Pharmaceuticals to develop protocols for Phase 2a trials and the documentation needed for regulatory agencies such as the Investigators Brochure.

Fogo Island’s Economic Ecosystem: Community Value-Chain, Labour Market, & Industry Diversification Analyses 2025

The Town of Fogo Island (TOFI) is a municipality formed in 2011 after the amalgamation of 11 small communities across an outport island on the Northeast coast of Newfoundland, which provides a unique set of challenges and special opportunities. Fogo Island’s economic development plan is to assess the local value-chains, labour market and industry ecosystem, to equip the Town of Fogo Island’s policymakers and local stakeholders with holistic economic information and insights on the community’s potential as a competitive player in the economic region. Focusing on the local labour market, community capacity, and the potential paths for economic diversification, this project aims to provide data-driven strategies to support community development, workforce development, and new industry development on Fogo Island.

Facteurs déterminants de la prise de risque : développement d’un outil de mesure

La pratique sportive, qu’elle soit compétitive ou récréative, bien que reconnue comme un facteur de protection pour la santé, comporte un risque de blessures ayant des conséquences significatives pour la santé des pratiquants. Des études montrent que les hommes de 12 à 24 ans (Hamel & Tremblay, 2012) sont les plus grandes victimes de traumatisme d’origine récréative et sportive (TORS). Une étude qualitative a permis de dresser un portrait des facteurs déterminants de la prise de risque en loisir. Un instrument de mesure de ces déterminants a été développé avec la consultation d’experts. Il importe maintenant de valider cet outil sur le terrain. Ce projet s’inscrit dans un partenariat avec l’entreprise Notyss et la Direction de la promotion de la sécurité afin de développer d’un questionnaire électronique convivial de l’instrument qui sera utile pour aider à développer des mesures de prévention des TORS plus efficaces

Non-invasive estimation of continuous arterial blood pressure using data from a wearable multi-sensor device.

Cardiovascular diseases are the leading cause of death worldwide, yet long-term continuous blood pressure monitoring, an accurate technique for prevention and early detection, remains largely unavailable outside of clinical settings. Such measurements presently require invasive, expensive, and impractical instrumentation thus preventing widespread use in non-clinical settings like at home and in research. This project aims to develop a machine learning-based algorithm that estimates continuous arterial blood pressure (cABP) from wireless wearable sensors, offering a non-invasive and accessible solution to this problem. By analyzing a unique dataset collected at Oslo University Hospital, the research will leverage multiple physiological signals, including photoplethysmography (PPG), electrocardiography (ECG), and skin temperature, to improve the accuracy and reliability of non-invasive cABP measurements in non-clinical settings. A collaboration between McGill University (Associate Prof. Georgios Mitsis, and Ph.D. candidate Rémi Dagenais) and Oslo University (Associate Prof. Ulysse Côté-Allard), this work will benefit both institutions by providing a novel method for estimating cABP non-invasively in cardiovascular and brain research while also advancing machine learning methods for multi-sensor data. Ultimately, this innovation could transform how we track blood pressure, leading to better cardiovascular health, earlier disease detection, and reduced healthcare costs.

Au-delà de Montréal : une nouvelle vision de l’accueil des demandeurs d’asile par la régionalisation.

Ce projet pilote innovant vise à repenser les stratégies d’accueil et d’intégration pour les demandeurs d’asile au Québec par la régionalisation. Contrairement à la centralisation des services d’accueil et d’intégration aux centres de Montréal, ce projet vise à mettre de l’accent l’intégration durable des demandeurs d’asile en région Cela contribuera à décongestionner Montréal et à revitaliser les communautés régionales. La stagiaire aura un rôle stratégique : elle sensibilisera les demandeurs d’asile, assurera leur accompagnement personnalisé et participera à l’évaluation de l’impact du programme. En intégrant des solutions organisationnelles novatrices et une analyse des résultats, le projet permettra d’optimiser les processus d’intégration, de mobiliser les parties prenantes et de maximiser les retombées sociales, économiques et démographiques du programme.

Efficient Computational Methods for Understanding Back Move-ment and Pain from Dynamic Data Modeling

This project uses machine learning algorithms to better understand back movement and low back pain. We apply supervised learning time series algorithms to data collected from Backtracks’ wearable de-vice — which consists of a malleable think curve that reads data collected from the participants’ spine movements. At each time step, such movements are represented as a curve; the dynamic evolution of this curve in time represents an individual’s spinal movements. The goal of this project is to use time series analysis to detect the type of activity (e.g., walking, sitting, bending forward) the participant is doing at any particular window of time. We will use time delay embeddings (Frank, Mannor et al. 2010) on individual observations from the Backtrack device in order to represent the dynamics and support vector ma-chines for the classification tasks.

Caractérisation de l’anisotropie d’alliages d’aluminium

L’aluminium, utilisé dans les secteurs des transports pour ses caractéristiques de légèreté et de robustesse, se trouve souvent sous forme d’extrusions comme les poutres de châssis et de carrosserie. Toutefois, les propriétés des pièces extrudées varient fortement en fonction des conditions de fabrication, telles que la température et la vitesse de déformation, ce qui entraîne une impossibilité de prédire les propriétés avant de les fabriquer. En particulier, une anisotropie peut être développée rendant difficile la prédiction et la standardisation des performances. Le projet proposé vise à alimenter des modèles prédictifs et des outils numériques pour mieux comprendre l’impact de ces paramètres sur la microstructure et les propriétés mécaniques des alliages d’aluminium, en étudiant des phénomènes tels que la recristallisation, la texture cristalline et la précipitation post-déformation. Cela permettra d’optimiser dès la conception les profilés extrudés, réduisant leur poids et améliorant leurs performances et durabilité, tout en offrant aux fabricants une meilleure maîtrise des processus de production, la possibilité d’adopter des formes plus complexes, et de réduire les coûts, dans le respect des exigences d’efficacité énergétique et de développement durable. Ce projet s’inscrit dans une lignée de projet et permet de compléter les campagnes de mesures faites précédemment.