19614 research outputs found
Sort by
Marginalized graduate students navigating the academy during the COVID-19 Pandemic: A phenomenological approach
Full text linkThe aim of this study was to understand marginalized graduate students’ use of academic libraries for research activities during the COVID-19 pandemic. Using a phenomenological approach, this study investigated the challenges, barriers, and coping strategies of marginalized graduate students from three Canadian universities. Focus groups were conducted to stimulate discussions and gather rich data from participants. Based on findings, this study offers several recommendations for inclusive spaces, accessibility across institutions, bridging divides, and more to address service gaps and improve library access for all users.Canadian Association of Research Libraries (CARL)FacultyReviewe
In vitro glioblastoma model on a plate for localized drug release study from a 3D-printed drug-eluted hydrogel mesh
No full textGlioblastoma multiforme (GBM) is an aggressive type of brain tumor that has limited treatment options. Current standard therapies, including surgery followed by radiotherapy and chemotherapy, are not very effective due to the rapid progression and recurrence of the tumor. Therefore, there is an urgent need for more effective treatments, such as combination therapy and localized drug delivery systems that can reduce systemic side effects. Recently, a handheld printer was developed that can deliver drugs directly to the tumor site. In this study, the feasibility of using this technology for localized co-delivery of temozolomide (TMZ) and deferiprone (DFP) to treat glioblastoma is showcased. A flexible drug-loaded mesh (GlioMesh) loaded with poly (lactic-co-glycolic acid) (PLGA) microparticles is printed, which shows the sustained release of both drugs for up to a month. The effectiveness of the printed drug-eluting mesh in terms of tumor toxicity and invasion inhibition is evaluated using a 3D micro-physiological system on a plate and the formation of GBM tumoroids within the microenvironment. The proposed in vitro model can identify the effective combination doses of TMZ and DFP in a sustained drug delivery platform. Additionally, our approach shows promise in GB therapy by enabling localized delivery of multiple drugs, preventing off-target cytotoxic effects.This research was funded Natural Sciences and Engineering Research Council of Canada (NSERC), Canada Foundation for Innovation (CFI) and BC Knowledge Development Fund.FacultyReviewe
Chemometric strategies for the detection of bromazolam and xylazine in illicit opioids using surface-enhanced Raman and infrared spectroscopy
Full text linkThe detection of trace adulterants in opioid samples is an important aspect of drug checking, a harm reduction measure that is required as a result of the variability and unpredictability of the illicit drug supply. While many analytical methods are suitable for such analysis, community-based approaches require techniques that are amenable to point-of-care applications with minimal sample preparation and automated analysis. We demonstrate that surface-enhanced Raman spectroscopy, combined with a random forest classifier, is able to detect the presence of two common sedatives, bromazolam (0.32--36% w/w) and xylazine (0.15--15% w/w), found in street opioid samples collected as a part of a community drug checking service. The Raman predictions, benchmarked against mass spectrometry results, exhibited high specificity for the compounds of interest (88% for bromazolam, 96% for xylazine) and sensitivity (88% for bromazolam, 92% for xylazine). We additionally provide evidence that this exceeds the performance of a more conventional approach using infrared spectral data acquired on the same samples. This demonstrates the feasibility of surface-enhanced Raman spectroscopy for point-of-care analysis of challenging multi-component samples containing trace adulterants.
Surface-enhanced Raman spectroscopy and infrared spectroscopy were integrated into two data fusion strategies - hybrid (concatenated spectra) and high level (fusion of high outputs from both models) - to enhance the predictive accuracy for xylazine detection. Three advanced chemometric approaches - random forest, support vector machine, and k-nearest neighbor algorithms - were employed and optimized using a 5-fold cross-validation grid search for both fusion strategies. Validation results identified the random forest classifier as the optimal model for both fusion strategies, achieving high sensitivity (88% for hybrid, 84% for high level) and specificity (88% for hybrid, 92% for high level). We demonstrate the enhanced practicality of the high level fusion approach, effectively leveraging the surface-enhanced Raman data with a 90% voting weight, without compromising prediction accuracy when combined with infrared spectral data. This highlights the viability of a multi-instrumental approach using data fusion and random forest classification to improve the detection of various components in complex opioid samples for community-based drug checking.Graduate2025-08-2
Reverse osmosis at the nanoscale: Investigating desalination membranes using classical molecular dynamics
Full text linkScarcity of freshwater is an ongoing crisis for a large portion of the global population, with developing countries affected the most. Reverse osmosis (RO) water filtration technology has played a critical role in increasing the availability of drinking water due to its ability to produce drinkable water from many different wastewater sources. One class of materials of growing interest in water filtration is covalent organic frameworks (COFs). COFs are advantageous as filtration materials due to their high crystallinity, stability, and tunable porosity. Recently, classical molecular dynamics (MD) simulations have gained traction as a tool to inspect water filtration processes at the nanoscale, particularly in applying new filtration materials like COFs. This work will investigate the microporous COF HPB-COF, possessing trigonal 12 Å pores. Quantities of interest, such as water flux and salt rejection will be derived from the generated MD trajectories. We will utilize large-scale simulations to investigate the relationship between desalination performance and membrane thickness, allowing a better understanding of how membrane dynamics change going from a monolayer to a bulk membrane.Jamie Cassels Undergraduate Research Awards (JCURA)UndergraduateReviewe
An Evidential Deep Learning Classifier with an Integrated Capability for Uncertainty Quantification
Full text linkWhile deep neural networks (DNNs) have demonstrated great proficiency in diverse tasks spanning various domains, the reliability of their predictions remains a subject of ongoing research. In the context of classification problems, there is a common misconception regarding probabilities generated by DNNs, falsely equating them with the confidence of the models in their assigned classes. Incorporating the softmax layer at the end of the network compels models to convert activations to probabilistic values between 0 and 1, irrespective of the underlying activation values. When activations are insufficient for accurate decision-making, raising uncertainty about the correct classification, a model should quantify its uncertainty about the true classification of the input data rather than making uncertain decisions. In this light, this study proposes a distance-based evidential deep learning (d-EDL) classifier with an additive capability for uncertainty quantification (UQ). The d-EDL classifier comprises two key components: the first utilizes convolutional neural network (CNN) layers for feature extraction, while the second incorporates designed layers for decision-making. In the second component, the first layer calculates basic probability assignments (BPAs) from the extracted feature vectors using a distance metric, measuring proximity between an input pattern and selected data representatives. A clustering algorithm is employed to form representatives for each data label; the closeness to a label representative reflects the potential belonging of the input to that label. The second and third layers employ combination rules to merge BPAs, leveraging probability theory and Dempster-Shafer (D-S) theory. The output of the d-EDL network is a probability distribution extended to include uncertainty as a class. An end-to-end training method is provided to train the proposed classifier, enabling joint learning and updating all network parameters. Five variants of the d-EDL classifier, each with a different number of data representatives, are trained on an image dataset, and their uncertainty quantification ability is assessed. The assessment involves evaluating the models in three scenarios, each with a common misclassification leading factor: noise, image rotation, and out-of-distribution (OOD) data. The results demonstrate the excellent capability of d-EDLs, especially those with 20 and 40 data representatives, to effectively quantify uncertainty rather than misclassification when faced with unfamiliar data.Graduat
Secure and Privacy-preserving Data Aggregation in Internet of Vehicles
Full text linkIn Internet of Vehicles (IoV), crucial data is aggregated to support the applications for automatic driving, intelligent transportation and smart cities. It is crucial to carefully address certain challenges in this process, particularly regarding security and privacy.
In this dissertation, we first target a representative IoV data aggregation scenario, fine-grained air quality monitoring. The major challenges we focus on include: a) the sensory data provided by vehicles usually vary in quality; b) there is a significant difference in traffic volumes of streets or blocks, which leads to a data sparsity problem; and c) the original sensory data, vehicle identities, and trajectories face risks of exposure. To address these issues, we propose a truth discovery algorithm incorporating multiple correlations, and extend it to a privacy-preserving framework, EAirQ.
EAirQ relies on a traditional end-to-end data aggregation architecture. Designing a new architecture specifically for vehicular networks may hold significant value. Thus, we introduce a privacy-preserving two-layered architecture with vehicle clusters. Instead of focusing on a specific application, we present how this architecture can be well adopted in a general distributed machine learning scenario. We named this part of the work CRS. CRS not only protects the local data, the identities and trajectories of vehicles, but also ensures the accuracy of aggregated learning models by handling packet loss in the application layer.
We further work on eliminating the limitations of the proposed two-layered architecture in the following three aspects: a) to provide fast and easy verification of messages within a cluster; b) to preserve vehicle privacy without adopting the pseudonym technique; c) to consider the adversarial behaviors of vehicles and enhance the security. Our solution introduces a novel concept, data approval, based on the Schnorr signature scheme. This part of the work, named SADA, meets more security requirements and is lightweight for vehicles.
In addition to exploring new solutions to preserve the privacy of vehicle identities and trajectories, we also pay attention to the latest industry standards. This part of the work focuses on tackling the challenge of certificate provisioning in the latest solution to satisfy the anonymous communication requirement in IoV. We propose a non-interactive approach, named NOINS, empowering vehicles to generate short-term key pairs and anonymous implicit certificates on their side. This new paradigm introduces the possibilities for many extensions and applications.Graduat
A global sensitivity analysis of parameter uncertainty in the CLASSIC model
Full text linkLand surface models (LSMs) have become indispensable for understanding the role of the terrestrial biosphere in the global climate system. However, the ability of LSMs to reproduce observed carbon, water, and energy fluxes varies considerably among models. Some of these deficiencies can be attributed to parameter uncertainties. Global sensitivity analysis (GSA) quantifies model output uncertainties caused by the uncertainty in model inputs. Our study conducts, for the very first time, a GSA for the Canadian Land Surface Scheme Including Biogeochemical Cycles (CLASSIC) model. Focusing on a site in the humid tropics, we evaluate the model's sensitivity for a wide range of ecosystem variables (17 in total). Considering a total of 90 parameters, we identify the top five most influential parameters using the qualitative Morris method per output variable. These influential parameters are then analysed using the quantitative Sobol' method. The analysis shows that the maximum carboxylation rate parameter has the greatest influence on almost all output variables considered. The impact of the maximum carboxylation rate is partially regulated by the canopy extinction coefficient's uncertainty. The results of this research will guide future efforts to optimize the model's performance more efficiently, focussing on a small subset of the 90 parameters.Raj Deepak S.N., Christian Seiler and Adam H. Monahan acknowledge the support of the Canadian Network for Research and Innovation in Machining Technology, Natural Sciences and Engineering Research Council of Canada (NSERC) [funding reference RGPIN-03787-2018].FacultyReviewe
Patterns of service utilization across the full continuum of care: Using patient journeys to assess disparities in access to health services
No full textHealthcare organizations have a contractual obligation to the public to address population-level inequities to health services access and shed light on them. Various studies have focused on achieving equitable access to healthcare services for vulnerable patients. However, these studies do not provide a nuanced perspective based on the local reality across the full continuum of care. In previous work, graph topology was used to provide visual depictions of the dynamics of patients’ movement across a complex healthcare system. Using patients’ encounters data represented as a graph, this study expands on previous work and proposes a methodology to identify and quantify cohort-specific disparities in accessing healthcare services across the continuum of care. The result has demonstrated that a more nuanced approach to assessing access-to-care disparity is doable using patients’ patterns of service utilization from a longitudinal cross-continuum healthcare dataset. The proposed method can be used as part of a toolkit to support healthcare organizations that wish to structure their services to provide better care to their vulnerable populations based on the local realities. This provides a first step in addressing inequities for vulnerable patients in accessing healthcare services. However, additional steps need to be considered to fully address these inequities.FacultyReviewe
University Newsletter, year 20XX
Full text linkAn award-winning work of fiction with the theme of equity, diversity, and human rights created by undergraduate student Ellen Pazder, selected by celebrity judge Thembelihle (Thembie) Moyo.2024 On the Verge Writing Contest second-place fiction winnerUndergraduat
BLOOD DONATION IN THE ERA OF BIOMEDICAL HIV PREVENTION AND GENDER-NEUTRAL DONOR SCREENING
Full text linkObjective.
Canada’s implementation of gender-neutral sexual behaviour screening allows sexually active men who have sex with men to donate blood for the first time. Public health campaigns promoting effectiveness of pre-exposure prophylaxis (PrEP) and undetectable = untransmittable (U=U) for HIV prevention heavily target sexual and gender minorities. Donor deferral policies remain in place for both methods. This thesis explores the tension between the effectiveness of these HIV prevention methods and donor policies considering them indicators of HIV risk.
Methods.
I wrote an algorithm approximating donor eligibility producing two analytic samples; one including PrEP use, one including HIV-negative men using U=U. I then estimate the proportion of donors who would be deferred for each prevention method. Chapter Two uses logistic regression to investigate PrEP use as a motivator for blood donation. Chapter Three describes HIV risk and protective factors for HIV and compares these observations to population health estimates of HIV incidence risk.
Results.
The algorithm identified n = 2,301 potential donors when PrEP users were included. Of these n = 85 (3.7%) would have been deferred for PrEP use. When repeated with HIV-negative donors using U=U, n= 2,354 donors were identified and n = 53 (2.3%) would have been deferred. PrEP use was not associated with willingness to donate. Estimates of HIV acquisition risk observed in the U=U analytic sample showed high risk of HIV acquisition. Contradictorily, a high number of combination HIV prevention strategies were also observed in the sample.
Conclusion.
It is likely donors are deferred solely for their choice of HIV prevention method. Having made a past donation was the best predictor of willingness to donate blood. Observed combination HIV prevention strategies employed by the U=U analytic sample did not support high public health estimates of HIV acquisition risk. Future research should explore PrEP adherence in samples of donors deferred for PrEP use and adjusting estimates of HIV acquisition risk to consider PrEP and U=U in risk estimates.Graduat