Innovative Projects Realized

Failure analysis on X-750 CANDU spacer material using bulk mechanical test and ion-irradiation

The current project will focus on understanding the behavior of one of the most important CANDU reactor components when it is subjected to the reactor environment. This study will develop a fundamental understanding of the X-750 material’s behavior resulting in innovative technologies that benefit the nuclear industry in Canada. In the short to medium term the work will support the life management and refurbishment of CANDU nuclear plants, assisting CANDU owners to protect their multi-billion dollar investment and sustain the provision of economical electrical energy for the benefit of the Canadian consumer. In the medium to long term these technologies will support the development of improved reactor components, for improved reactor designs. This will help to assure future CANDU sales domestically and abroad and support the ongoing Canadian nuclear industry.

Population pharmacokinetic-pharmacodynamic modeling of mycophenolate and tacrolimus in pediatric kidney transplant recipients

Mycophenolic acid (MPA) and tacrolimus (TAC) are approved for use in post-transplant immunosuppressants to prevent graft rejection after solid organ transplantation. MPA and TAC are frequently used in combination as an immunosuppressive regimen in pediatric kidney transplant recipients. However, these two drugs have narrow therapeutic ranges and large interindividual pharmacokinetic and pharmacodynamic variabilities. What’s more, in recent years, some previous studies have reported that there were drug-drug interactions (DDI) between TAC and MPA. Therefore, a new study is needed to build a population pharmacokinetic-pharmacodynamic model for TAC and MPA in pediatric kidney transplant recipients. The overall goal is to establish a population structural model incorporating both drugs and identify clinically significant covariates. Evaluation of the influence of drug-drug interactions and individual variabilities on the PK-PD changes will provide striking improvement in post-transplant immunosuppressive outcomes in pediatric patients treated with TAC and MPA.

Building Information Modelling (BIM) for First Nations Land Planning andInfrastructure Management — Kitigan Zibi Anishinabeg

Working in partnership, Kitigan Zibi Anishinabeg (KZA) and the Carleton Immersive Media

Studio (CIMS) propose the development of a novel, hybrid digital platform that brings

together the progressive community development initiatives of KZA with CIMS’ established

research record in the field of building information modelling (BIM). The research

challenges involved in the development of a comprehensive BIM for KZA are considerable.

First, the physical size of the property, some 18000ha, is well beyond the conventional

scale of a BIM. Secondly, the complexity and scale of the assets that KZA plan to

incorporate through the BIM (community archival information, data related to infrastructure,

existing GIS, planning proposals) challenge the parameters of current best practice for

BIM. We propose a hybrid BIM/GIS (geographic information system) that integrates new

and existing 2 and 3 dimensional data through an interoperable viewer client.

Mimicking Lung- and Gut-Pathogen Interactions Using a Hydrogel-Based Mucus Layer

The objective of this research is to investigate long-term interaction between host cell and pathogen. Research will be proceed using ATPS(Aqueous two phase system) and hydrogel. ATPS enable pathogen to form localized biofilm on mammalian cell tissue. To see long-term interaction, biofilm will be stably localized on mammalian cell using hydrogel, which protect mammalian cell from toxicity of PEG(Polyethylene glycol) and also form mucus. Interaction of Long-term biofilm – mammalian cell interaction will be analyzed by various methods such as cytokine response, cell viability, cell morphology using electro microscope, etc.
This research goes further from previous in terms of pathogen specificity, and cell line specificity, and adapting mucus using hydrogel. Thus, it can say result of research will reveal more realistic interaction of host-microbes.
For further study pathogen biofilm removal research using predatory bacteria will be done. Previous research shows predatory bacteria have biofilm removal effect of biofilm. This study will investigate whether predatory bacteria remove pathogen biofilm and decrease of pathogen biofilm virulence.

Utilization of biochar amendment for reduction of dissolved organic carbon in runoff water from soils covered with plant residue

Surface cover of soils with plant residue such as straw is known to be an efficient method to reduce soil erosion and nutrient loss from soils. However, from my previous research, it was revealed that surface cover reduced soil erosion and loss of nitrogen and phosphorous but increased loss of dissolved organic carbon (DOC) due to decomposition of the plant residue. As loss of DOC not only hampers soil conditions but also causes water pollution, it is necessary to develop a method to reduce DOC loss from the soils with surface cover. In this project, I will use biochar, which has a great capacity to sorb and immobilize DOC, as a soil amendment in combination with surface cover to reduce DOC loss. For the objective, I will conduct a series of experiments that include determination of the capacity of biochar to sorb DOC and investigation of the changes in DOC concentration in the soils amended with biochar in the presence and absence of plant material (surface cover) in the lab. I expect that the results of this research should contribute to increase in the efficiency of surface cover to reduce soil, carbon, and nutrient loss from soils.

The Next Generation Agriculture: Role of Functional Microbiome in Cannabis Breeding strategies against biotic stress

Plants including Cannabis host distinct beneficial microbial communities on and inside their tissues designated the plant microbiota from the moment that they are planted into the soil as seed. Understanding the microbial partnerships with Cannabis has the potential to affect agricultural practices by improving plant fitness and production yield of Cannabinoids. Much less is known about these beneficial Cannabis-microbe interactions, particularly,the role that Cannabis may play in supporting or enhancing them. This proposal aims to characterize the bacterial diversity, associated with susceptible and resistant Cannabis varieties to grey mold (GM) and Powdery mildew (PM) on Cannabis. We hypothesize that different varieties (susceptible and resistant) recruit and maintain different microbial communities and that select microbial strains are able to suppress GM and PM infections and increase protection. We also intend to compare the efficacy of CELEXT07, a new botanical product developed by the industrial partner proven effective against both diseases in greenhouse trials alone or with combination of the best endophytic microbes.

Development of machine learning and artificial intelligence based tools to improve efficiency in financial services – Year two

Our interactions with actors in the financial services industry, including our partner company, uncovered that they possess large amounts of data pertaining to investors and markets, but have yet to extract/learn information of significant value from that data such as expected actions by clients.
The industry is conscious of this, but while they are making the needful investments in IT, they report lack of academic expertise in machine learning (ML) / artificial intelligence (AI) to unlock full potentials of such investments. This project will combine academic and industrial expertise to resolve this bottleneck. We will develop ML / AI based tool to allow predictions of actions by client, specifically client churn and to help identify optimal fee structures as well as targeted populations, which Purefacts views as necessary to improve productivity and earning potential. We will also develop descriptors of accuracy of such predictions.
The feasibility of the project is assured by deep expertise of each party in respective domains: this applicant’s in applied math, coding and ML, academic supervisor’s in ML methodologies, and Purefacts’ expertise in financial services to individuals and major financial institutions.
Methods and tools developed in the project will be applicable to other industries.

Development of machine learning and artificial intelligence based tools to improve efficiency in financial services

Our interactions with actors in the financial services industry, including our partner company, uncovered that they possess large amounts of data pertaining to investors and markets, but have yet to extract/learn information of significant value from that data such as expected actions by clients.
The industry is conscious of this, but while they are making the needful investments in IT, they report lack of academic expertise in machine learning (ML) / artificial intelligence (AI) to unlock full potentials of such investments. This project will combine academic and industrial expertise to resolve this bottleneck. We will develop ML / AI based tool to allow predictions of actions by client, specifically client churn and to help identify optimal fee structures as well as targeted populations, which Purefacts views as necessary to improve productivity and earning potential. We will also develop descriptors of accuracy of such predictions.
The feasibility of the project is assured by deep expertise of each party in respective domains: this applicant’s in applied math, coding and ML, academic supervisor’s in ML methodologies, and Purefacts’ expertise in financial services to individuals and major financial institutions.
Methods and tools developed in the project will be applicable to other industries.

Machine Learning to Predict Temporomandibular Disorders Risk from Genotypes

The goal of this project is to develop new machine learning methods and computational strategies to mega-analyze data from well-characterized datasets on chronic pain conditions to develop a genetic predictive tool. This tool will be implemented in an online interactive dashboard and used by the Quebec Pain Research Network (QPRN) community. This collaboration with Plotly will make the developed machine learning models more accessible to applied researchers by: 1) visualizing the genetic effects which drive the predictions, 2) allowing users to interactively generate new predictions over a range of parameters and visually compare the outputs, and, 3) producing different graphics of the data to reveal details that might be hidden by summary statistics.

Idiomatic foreign function interface generation for user-specified target languages

For different software packages created using different tools to interoperate, an intermediate layer called API bindings is needed. These bindings can be created by hand, but that takes time and needs to be updated whenever one of the packages changes. Since these bindings are often quite repetitive, it is reasonable to try and generate them automatically, saving time both creating them in the first place and updating them due to changes.
There are existing tools that allow different sorts of automation in generating bindings, but these tools often make strong assumptions about what the result should look like. These results can require adapting by hand, which can be as much of a time sink as writing the bindings manually. We propose a more flexible way of generating these bindings, which aims to save time for PDFTron employees by automating more of this process for them.

Etude et optimisation des operations de mise en forme courbee de conduit

Le stagiaire devra ultiliser ses connaissances acquises dans le domaine des composites pour resoudre une problematique de l’entreprise FRE Composites (2005) inc. Celle-ci porte sur le pliage de leurs conduits de composites. La problematique et que lors pliage, il y a parfois creation de rupture du materiau et/ou de flambage local, et que le conduit a tendance a perdre son angle et rayon initial avec le temps. Le projet prend en compte les differentes phases de polymerisation du composit lors des etapes de fabrication (enroulement filamentaire, cuisson, refroidissement, prechauffage, pliage, refroidissement, post-pliage) etplus partciulierement le comportement mecanique lors de la phase de pliage. L’objectif intermediare du projet est donc de comprendre et d’identifier les parametres critiques pour l’operation de pliage d’un thermodurcissable. L’objectif final est d’implanter une nouvelle facon de plier des conduis en materiaux composites permettant d’eliminer ou de reduire les rejets, en plus d’optimiser les proprietes mecaniques……..

Development and validation of an automated diagnostic tool for wound imaging – Year two

Over 6.5 million people in North America live with chronic wounds which pose a burden on their quality of life and the healthcare system. Chronic wounds are estimated to cost over $30 billion per year. Swift Medical is a pioneer in point-of-care imaging for wounds. Their mobile apps allow the reliable and accurate measurement of wound characteristics, making it an ideal tool to track healing and identify healing patterns. Using artificial intelligence/machine learning and a large database of wound data that Swift Medical uniquely possess, we propose the development of a diagnostic tool to classify wound images and its validation in an independent cohort composed of patients receiving care by Professor Gregory Berry at the McGill University Health Center. The resulting algorithms will be used by Swift to enhance the capabilities for their mobile technology, which could improve patient care by monitoring patients at high-risk of chronic wounds, such as people with diabetes or impaired mobility, promote widespread access to telemedicine in remote communities, and reduce the overall cost of chronic wound treatment to the healthcare system.