Projets novateurs réalisés

Machine Learning and Interactive Visualization to Detect Dose Associated Adverse Drug Events in Drugs Prescribed to Older Adults

In this research project, we aim to address the uncertainties in the efficacy, safety, and dosing of prescription drugs, particularly for older adults who are more likely to take multiple medications. Pre-market drug trials often exclude the older population, leading to potential risks. Our proposed method involves utilizing machine learning and interactive visualization techniques to analyze healthcare databases in Ontario. By simultaneously examining hundreds of drugs in older adults and comparing high-dose versus low-dose effects, we aim to identify potential adverse reactions. By replicating our findings in other regions and translating them into guidelines, we will enhance prescription drug safety and improve patient outcomes.

Les crises européennes : Un moment clé dans la formation de l’identité européenne

Les récents changements institutionnels de l’Union européenne, tels que le Traité de Lisbonne, ont renforcé son rôle et son impact, tandis que les crises, comme le Brexit et la crise migratoire, ont touché divers domaines politiques et la vie quotidienne des Européens.

Ce projet explore comment les crises ont pu avoir un impact sur les mécanismes d’identification qui pourrait avoir eu un impact sur les identités individuelles et collectives envers l’UE. Ce projet vise à comprendre l’impact des crises sur la formation de l’identité européenne en examinant quatre mécanismes clés : l’exposition aux institutions européennes, la solidarité entre Européens, la perception de problèmes comme de nature « européens » et la création d’une culture mémorielle européenne.

Ce projet de recherche utilisera la méthode des focus groups pour étudier l’identification à l’UE et ils seront réalisés à Louvain-La-Neuve, en mettant l’accent sur les expériences vécues pendant les crises récentes.

Simulation d’insertion de cathéters veineux périphériques par écho-guidage

L’insertion de cathéters veineux périphériques permet l’injection de médication à un patient. Cette procédure est
réalisée chez environ 70% des patients admis dans un hôpital canadien. L’insertion du cathéter peut être difficile,
particulièrement chez les patients pour lesquels les vaisseaux sanguins ne sont pas visibles à la surface de la
peau. De plus, cette procédure comporte des risques non négligeables de complications en cas d’erreur ou de
tentatives multiples d’insertion. Le recours à l’imagerie par échographie pour guider l’insertion permet de réduire
les risques. Cependant, cette technique est difficile à maîtriser. L’entreprise Sononurse a conçu un dispositif qui
permet de faciliter cette procédure, ce qui a le potentiel d’élargir l’adoption de la technique et ainsi de réduire le
taux de complication associé. Le projet proposé vise à évaluer l’utilité du dispositif de Sononurse en le comparant
à la méthode classique dans un simulateur en réalité virtuelle.

Analyzing the Inequity of Home Equity Investments

In the emerging sector of fintech, a novel concept known as a home equity investment is gaining traction, offering homeowners a way to access their property’s value without selling it. This financial instrument involves an investor providing a cash infusion to homeowners in exchange for a 20% ownership stake in their home. Homeowners then have a decade to settle the investment, either by selling their property or repaying the investor based on its appraised value. While four North American companies offer this service, academic research on them is lacking, leaving the economics, incentives, and consequences unexplored. Our research aims to address this gap by constructing a comprehensive econometric model of the housing market, considering both homeowners and investors. We focus on modeling homeowners’ “impatience” (i.e., discount factor), aiming to determine the impatience levels that typically lead individuals to opt for this financial tool. We hypothesize that high impatience, often linked to urgent financial needs such as medical expenses or education costs, drives homeowners towards these investments. Despite appearing helpful, these investments may potentially exploit vulnerable situations. This project seeks to shed light on this situation.

How does the brain generalize learned event-pain relationships to novel contexts?

Pain is a complex unpleasant experience that varies from a person to another. Learning, priors and expectations are known to have a substantial influence on subjective pain experience. For example, having expectations or being told that an event will be painful can be sufficient to shift a person’s pain experience into a more painful realm. When such modulation of pain occurs, changes of pain-processing systems in the spinal cord and in the brain can be observed using neuroimaging methods. This project uses functional Magnetic Resonance to unveil how the brain can learn to associate some situations with high pain outcomes while discriminating other situations to be associated with lower painful outcomes. The use of innovative AI-based statistical algorithms (such as machine learning and predictive models) will be at the heart of the project. We expect to find which brain regions are involved in producing a different pain experience in response to different conditions of learning.

Development of Microneedle-Based Biosensing Systems

Microneedles are small sub-millimeter projections that allow for pain free access to the dermal layers of the skin. This allows for enhanced drug delivery and biosensing capabilities that are much more accessible and comfortable for the patient. This work aims to improve the capabilities of a solid metal microneedle system designed for use in electrical bio signal monitoring to enhance accuracy of the system while maintaining the simplicity of the current patch style electrodes used for electrocardiography (ECG), electromyography (EMG), etc. By utilizing an existing fabrication process for these microneedles, this work will study the impacts of array design, fabrication techniques, and surface modifications to improve system accuracy, leading to better patient outcomes, higher manufacturability, and increased reliability. Specifically this work will focus on the development of an electrode system for EMG analyses of patients with Parkinson’s disease.

Outcome and Resource Use of Patients with Primary Liver Cancer

Primary liver cancer (PLC) is one of the fastest growing cancer in Canada and is a heavy global burden. To efficiently allocate the health care resource to the patients with PLC, outcome and cost analysis of the patients with PLC may be a right direction. In order to obtain enough information to accurately assess the status of patients with PLC, this proposed study will take advantages of Electronic Health Record (EHR) data. EHR data is the original health care records of detailed information of patients during their hospitalization including not only the administrative information but also laboratory and radiological test results, death record, detailed cost information, detailed present and past illness, pathology diagnosis, records of use of medication etc. This work is unique and challenging because of extracting clinical data from EHR. This work is not only building foundation for analyzing large EHR and also generates robust models for predicting outcomes.

Design solutions for larger-diameter spoolable coating composite piping identifying opportunities and limitations

This project is motivated by a significant demand for eco-friendly and recyclable composite structures for lightweight spoolable composite piping for oil, gas and water applications. Therefore, this partnership (with Flexpipe at Mattr Infrastructure Technologies) aims to determine an application envelope for multilayer spoolable larger-diameter composite piping by considering lightweight fiber-reinforced thermoplastic composite structures. In addition, this project is crucial for the partner and Canada because: i) there is a growing need to improve the performance of composite structures that are recyclable and environmentally friendly in various sectors, such as oil and gas; ii) there is currently no clear understanding of the long-term performance of commercially available fiber-reinforced thermoplastic structures for the partner-specific applications, i.e., linepipe for oil/gas and water applications; iii) there is a scientific opportunity to contribute multi-scale testing and characterization of damage mechanisms due to aging and environmental exposure to the limited technical literature; iv) there are opportunities for knowledge transfer to other Canadian organizations dealing with issues in technology comparable to those faced by the partner, i.e., the use of durable, lightweight materials in harsh environments.

Adaptive instructional system for safe marine operations in icy waters

Seafarers need ice navigation training to protect the safety of people, the environment, and ships while driving in ice covered waters. In this project, Virtual Marine’s simulators will be used to create a new method of delivering ice navigation training – an adaptive instructional system. In an adaptive instructional system, a computer develops a personalized training experience for each student based on what they know, how they learn, and what skills they have. The goal of developing this adaptive instructional system is to make training more effective for individuals and attractive to potential customers, which could benefit Virtual Marine by improving its training programs and increasing its market share.

L’évolution des groupes parlementaires à l’Assemblée nationale : des motivations individuelles aux effets institutionnels

Le projet de recherche explore l’augmentation du nombre de groupes parlementaires à l’Assemblée nationale française, passant de quatre en 2007 à dix en 2022, en examinant les facteurs qui expliquent cette évolution. Il examine si les modifications des règles de procédures de l’Assemblée, et notamment celles relatives aux groupes parlementaires, ont contribué à cette croissance. Le projet prend en compte le fait que la création de nouveaux groupes parlementaires implique des coûts pour leurs fondateurs et étudie l’impact des changements de règles sur ces coûts. L’analyse couvre la période de 1910 à 2022 et inclut des entretiens avec des députés et des assistants parlementaires des 15 dernières années pour évaluer l’influence des changements de règles sur l’augmentation des groupes parlementaires.

LANGA – Big Data for Language Learning

LANGA is a project at Copernicus Studios (Halifax, NS) that is experimenting with the use of automated speech recognition in on-line video games to assist in language learning. Over the past 2 years, the company has been developing prototypes of the game and testing them in conjunction with the NeuroCognitive Imaging Lab at Dalhousie University. A key element to both the commercial success of the product and the research supporting its efficacy is scaling from lab-based testing of small numbers of people, to on-line delivery that supports thousands and possibly millions of users. In scaling to large numbers of users, we need to maintain the ability to test changes to the games and rapidly and accurately visualize the results of these tests. To enable this, we require the development of a robust and scalable database architecture (“back end”) to store user profiles and gameplay data. This database will need to be designed around both the needs of the game itself (e.g., tracking users and their progress), the unique demands of the speech recognition technology (storing sound files and speech recognition data).

Community Integration

Community Integration is a key indicator of long-term stability and well-being. However, few people with experiences of homelessness experience belonging within a community. This research project aims to discover the reasons for this disconnection through facilitated focus groups with individuals with lived experience and offer new opportunities for exploring new ways of achieving community integration. The objective of this project is to identify the expectations and needs of individuals in the HomeBase Program at The Alex to refresh and refine community integration intervention and improve outcomes for homeless or recently housed individuals.