4 Mitacs Globalink (GRI) research projects for Summer 2027.
1. Care trajectories in mental health: modelling and prediction
Psychotic spectrum disorders (PSDs), including schizophrenia, schizoaffective disorder, and bipolar disorder, affect approximately 3% of the population. This population have complex care needs and experiences frequent transitions of care between the different settings they consult. They are also associated with considerable healthcare use, negative outcomes (e.g., suicide, avoidable hospitalizations), and reduced life expectancy. Thus, these are costly health conditions from societal and health system perspectives, despite the availability of effective treatments. Care trajectories (CTs), i.e., the pathways of health care use across the lifespan, remain poorly understood. Furthermore, little is known about the impact of care transitions, e.g. between family medicine and psychiatry, or other contexts leading to loss of follow-up or non-adherence to care. Adequately modeling those CTs (biostatistical or artificial intelligence models) could help improve early identification of at-risk individuals, guide preventive strategies, and ultimately reduce the burden of untreated psychosis through timely, targeted interventions.
The aim of this internship is to analyze CTs across PSD stages to identify opportunities for better screening, early care, and management of established disease.
Research area, student roles & skills
Research area: I am a professor at Université de Sherbrooke (Faculté de médecine et des sciences de la santé). My main areas of expertise are biostatistics, epidemiology, and artificial intelligence.
I am interested in statistical and machine learning methods for modeling and predicting outcomes associated with complex health issues, using large clinical administrative databases. My research focuses, among other things, on high use of health services, chronic diseases, and environmental health in the context of digital health.
Student roles:
Overall project management: the intern will be expected to demonstrate leadership and autonomy to manage a clearly defined research project, while also contributing their methodological or clinical expertise to the project.
CTs modeling: the intern will work with a mental health database and composite variables. They will model CTs using statistical or artificial intelligence models. Our team is currently leading a scoping review to determine the most appropriate models. The intern must be able to manipulate, analyze, and interpret quantitative data using descriptive and inferential statistics.
Outcome prediction: the intern will use predictive models (e.g., logistic regression or random forests) to predict the most significant outcomes associated with CTs. They will need to select the most relevant explanatory variables and appropriate models. Examples include a first hospitalization, a prescription of inappropriate medication, or death.
In the analyses, a methodological profile will enable the implementation of statistical/AI models suited to the available data and the research question, while a clinical profile will allow for contextualizing the statistical analyses, interpreting care models, and grounding the results in clinical practice.
Manuscript writing: the intern will be included in manuscript writing and, if interested, will be included in the submission process.
Food for thought: the intern will be expected to participate in the lab thinking and discussions. Their ideas and insights will be highly valued.
Skills required:
Skills:
- Background in biostatistics, pharmacoepidemiology, applied mathematics, or a related field with a focus on quantitative analyses (intermediate level; advanced level is a plus)
- Critical thinking and curiosity
- Presentation of results and communication
- Clinical experience or working experience with medications, NHP, or large health databases is a plus
Technological aspects:
- Proficiency in statistical analysis software (e.g., Excel, R, Python, SAS)
- Comfortable using the Office suite for collaborative work (e.g., Teams, Word, SharePoint)
2. Classification of natural health products and description of their use in middle-aged and older Canadians: a CLSA study using machine learning
Even though Canadians spent an estimated 3.6 billion dollars on natural health products (NHP) in 2005, limited and outdated information is available on which products are used, for what reasons, and how much they cost. Such information is necessary to conduct pharmacoepidemiological studies on the actual impact on health of these products and their interactions with medications.
1) To address those gaps, we propose, as a first step, to apply machine learning methods to develop a classification of NHPs. This classification will be based on free-text data from Health Canada Licensed Natural Health Products Database. It will extend the anatomical, therapeutic, and chemical classification of medications to NHPs.
2) Next, using the developed classification, we will analyze data from the Canadian Longitudinal Study on Aging to describe the baseline trend and six-year evolution of NHP use and associated costs, as well as factors related to their use. Relationships between NHPs used with medications and diabetes will be investigated as a case study using regression models.
Research area, student roles & skills
Research area: I am a professor at Université de Sherbrooke (Faculté de médecine et des sciences de la santé). My main areas of expertise are biostatistics, epidemiology, and artificial intelligence.
I am interested in statistical and machine learning methods for modeling and predicting outcomes associated with complex health issues, using large clinical administrative databases. My research focuses, among other things, on high use of health services, chronic diseases, and environmental health in the context of digital health.
Student roles:
Overall project management: the intern will be expected to demonstrate leadership and autonomy to manage a clearly defined research project, while also contributing their clinical expertise to the project.
Development of the classification: the intern, under the guidance of the principal investigators, will work with a database of PSNs and will help create a classification of the main products. This step will require database management and variable matching.
Data analysis: the intern must be able to handle, analyze, and interpret quantitative health data using descriptive and inferential statistics. A methodological profile will enable the implementation of statistical and AI models tailored to the available data and the research question, while a clinical profile will allow for the contextualization of statistical analyses, the interpretation of care models, and the application of results to clinical practice.
Manuscript writing: the student will be included in manuscript writing and, if interested, will be included in the submission process.
Food for thought: the student will be expected to participate in the lab thinking and discussions. Their ideas and insights will be highly valued.
Skills required:
Skills:
- Background in biostatistics, pharmacoepidemiology, applied mathematics, or a related field with a focus on quantitative analyses (intermediate level; advanced level is a plus)
- Critical thinking and curiosity
- Presentation of results and communication
- Clinical experience or working experience with medications, NHP, or large health databases is a plus
Technological aspects:
- Proficiency in statistical analysis software (e.g., Excel, R, Python, SAS)
- Comfortable using the Office suite for collaborative work (e.g., Teams, Word, SharePoint)
3. Exploring Local Climate Change Action in Canada: Place-based strategies, community engagement, and environmental justice
To establish common objectives and help coordinate international action on climate change, the Paris Agreement was adopted by 196 parties at the UN Climate Change Conference in 2015. Its over-arching goal was to limit temperature rise to between 1.5 and 2 degrees Celsius, in order to avoid the most catastrophic effects of climate
breakdown. Meeting this goal included mandates to reduce greenhouse gas emissions by 45% by 2030 and to reach net zero emissions by 2050. While such coordinated efforts at national scales are important, lower levels of government are more often responsible for developing and implementing policies and practices designed to mitigate and adapt to climate change. The mandate is especially clear for cities to reduce emissions – where nearly 75% of all emissions originate, and where most (approximately 80%) of economic activity is situated. Within this context, there is the need for more proactive efforts by local governments to develop and implement locally-relevant climate change policies. While emerging evidence suggests a move towards local action (such as 2400+ municipalities and other governments declaring climate emergencies), several key gaps in knowledge exist. These include understanding what policies and actions local governments are prioritizing, how communities (i.e., stakeholders and residents) are being engaged for 'buy-in' on climate strategies, the level of civic participation in the planning and policy development process, and whether climate adaptation and mitigation strategies are inclusive, equitable, and just. To address these gaps, this project is framed around three key research objectives:
• RO#1: identify and analyze climate change policies and strategies are being implemented by
local governments in Canada and the UK;
• RO#2: explore processes facilitating community engagement and action, their inclusiveness, and efficacy;
• RO#3: investigate the relationship between government and residents, to better understand civic participation.
Research area, student roles & skills
Research area: Dr. Chad Walker is an interdisciplinary environmental social scientist with particular research interests around justice, equity, and public support for low-carbon transitions. Recent published research includes studying the impact of environmental justice in shaping support for wind energy, critically investigating the meaning of community energy, and using diverse methodologies to better understand pathways for Indigenous-led renewable energy development. He has been fortunate to publish in a variety of high-impact journals spanning several disciplinary boundaries, including: Energy Policy, Environment and Planning A, Environmental Policy and Planning, and Energy Research and Social Science.
Student roles:
Depending on qualifications, the student(s) may be responsible for data collection, analysis, and/or writing.
Skills required:
Interest and familarity with:
a) climate change and clean energy solutions
b) social science research (qualitative and/or quantitative approaches)
4. Systèmes d'aide à la décision pour l'industrie 4.0
Le candidat sera intégré à une équipe de recherche qui développe des modèles et logiciels d'aide à la décision dans le domaine de l’industrie 4.0. Différentes techniques d'intelligence artificielle (Mixed-Initiative Systems, Constraint programming, Machine Learning) sont utilisées pour rendre les logiciels plus intelligents. Différents outils sont utilisés tels que la réalité augmentée où à l’aide de systèmes de visions avancées, l’opérateur voit en temps réel le positionnement de pièces, ou encore un robot capable de faire du suivi en temps réel d’un objet, de l’usinage de précision ainsi que de la manutention. Tous ces outils seront mis à la disposition de l’étudiant dans le cadre de son stage pour développer la partie algorithmique en arrière d’un processus industriel 4.0.
Research area, student roles & skills
Research area: La recherche au Lab-Usine se situe au niveau de l'ingénierie automatisée à base d'intelligence artificielle et de modèles mathématiques, à des systèmes de production auto-reconfigurables avec lancement automatique de la production (planification automatisée des opérations en temps réel) et suivi de la production s'appuyant sur la science des données, l'internet des objets, la réalité augmentée et les jumeaux numériques.
Student roles:
Le candidat sera familiarisé à ces techniques d'intelligence artificielle et aura l'occasion des les programmer dans ces outils.
Skills required:
Études en informatique ou génie informatique ou génie logiciel ou génie industriel Analyse/programmation orientée-objet. Programmation en C#.net ou équivalent ou Python. Intérêt pour la recherche et l'intelligence artificielle