Carleton University Institutional Repository
Not a member yet
20816 research outputs found
Sort by
Investigation of Machine-Learning Techniques for Pulmonary Artery Pressure Estimation from Electrical Impedance Tomography Images
No full textPulmonary artery pressure (PAP) is measured for the diagnosis and monitoring of pathologies such as congestive heart failure. Unfortunately, current modalities used for PAP measurements have disadvantages which reduce their use. One proposal to address these issues is the measurement of PAP using a continuous and non-invasive imaging modality - electrical impedance tomography (EIT). Previous work used a model-based algorithm relying on the known association between thoracic conductivity changes and PAP. This thesis attempted to improve these results by predicting PAP using neural networks (NN). Multiple NN architectures were trained and tested on a dataset of eight cardiac failure patients, and their prediction methodologies were analyzed. The resulting NN models performed well on seen patients, but failed to generalize well on unseen patients. The results present the possibility that NNs may be able to predict PAP given a more expansive dataset
Genomics Approaches to Improve the Recovery of Bacterial Pathogens in Foods
No full textFoods are important vehicles for transmission of infectious bacterial pathogens. Reliable methods for detecting these organisms would ultimately result in fewer contaminated products released into the marketplace. Leveraging genomics tools such as whole genome sequencing (WGS) and metagenomics within the domain of food microbiology presents promising opportunities for enhancing diagnostic methods. This thesis demonstrates the value of genomics-based approaches for improving microbiological methods for foodborne pathogen detection. Two examples of recovery methods developed using WGS-informed antimicrobial resistance (AMR) traits are presented. Furthermore, the application of metabarcoding analysis investigating microbial compositions in food enrichments through 16s rDNA sequencing, is explored to gain a deeper understanding of pathogen growth relative to the background microbiota of food samples. The first case exemplifies a custom (strain-specific) selective enrichment approach for Shigella recovery from outbreak-associated food. This method incorporates enrichment media supplemented with antibiotics chosen based on the WGS-predicted AMR features of the target pathogen. Chapter 2 evaluated the feasibility of the antibiotic-supplemented custom enrichment media in enhancing the recovery of viable drug-resistant Shigella during competition with interfering microorganisms. Chapter 3 further investigated the performance of the custom selective enrichment media for Shigella recovery from baby carrots linked to historical shigellosis outbreak. The addition of the appropriate antibiotics in food enrichment media reduced the relative proportion of competing bacteria in the enrichment cultures and significantly enhanced the recovery of drug-resistant S. sonnei. This demonstrates the potential of genomically-informed selective enrichment in aiding foodborne shigellosis outbreak investigations. To explore whether the application of the genomcally-informed selective enrichment approach could be expanded for an entire pathogen species, the second example (chapter 4) presents an overview of a novel Salmonella selective enrichment broth (Minimal Salts Medium supplemented with amikacin) developed based on a species-specific aminoglycoside resistance genotype (aac(6’)-Iy or aac(6’)-Iaa). The performance of this medium relative to the current enrichment methods by Health Canada was evaluated using chicken feed contaminated with S. Enteritidis. While the novel approach did not provide improved selectivity relative to current methods, 16S rDNA sequencing provided insight into enrichment dynamics in the media evaluated that could be applied to further refinement of this methodology
Investigation of Biotransformation of 6:6 PFPiA into C6 PFPA during Anaerobic Digestion of Sludge
No full textThe biotransformation of perfluorinated phosphinic and phosphonic acids (PFPiAs & PFPAs) have been studied in animal and plant models, but not in wastewater treatment plant sludge. This is concerning since the toxicity and general research into PFPiAs and PFPAs is lacking and biosolids, a by-product of sludge treatment, is used as fertilizer for commercial crops. The purpose of this experiment is to fill in this void in research by investigating whether the microbes in sludge can bio-transform 6:6 PFPiA into C6 PFPA during a 30-day anaerobic sludge digestion. Four samples were prepared to measure the microbial, non-biological, and abiotic degradation of 6:6 PFPiA in comparison to a sterile control sample. Over the 30-day digestion period, it was found that 6:6 PFPiA levels decreased while C6 PFPA levels increased in the microbial sample. It was not determined if this degradation was caused by the cleavage of 6:6 PFPiA’s carbon-phosphorus bond
The Scale of Industry: Visceral Understanding of Shipping and the St. Lawrence Seaway
No full textHow do we engage with a world that operates on a larger-than-life scale? How can we communicate the scope of systems of such a magnitude that it begins to lose meaning and connection to a person? In the conveyance of mass statistics and data, an embodied, tangible knowledge may be lost. The portion that feels “real”. Yet these systems have an integral effect on our lives. Understanding them, especially as a lay person, can be an important influence in determining ones political, social, and personal values. Often, environments are used to reintroduce weight to abstracted concepts. The videos of endless garbage patches floating over the ocean, grey fields of the burnt remnants of a forest fire, contaminated water brown with industrial waste. All environments, not just visual but spatial. Does architecture, a field primarily concerned with the communication of space, provide us the opportunity to convey vast and inhuman topics
Effects of Interactive Engagement Approaches on Mind-Wandering, Task Load, and Usability in Educational Videos
No full textEducational videos are widely used, but their effectiveness is often limited by mind-wandering. It is not clear whether interactions like "click to continue" buttons can reduce mind-wandering and enhance learning. The locations of the buttons (static vs. dynamic) and the input type (mouse clicks vs. gaze input) may play a role in how effective such interactive elements are. This thesis evaluated four types of buttons – Static-Click, Dynamic-Click, Static-Gaze, and Dynamic-Gaze – based on how they affect mind-wandering, learning, task load, and usability in educational videos. User preferences were also collected. Results showed no impact of these interactions on mind-wandering or learning. Gaze-based interactions were easier and faster than mouse-based ones and elicited lower physical load. Most participants preferred Static-Gaze but found Dynamic-Click most effective in reducing mind-wandering. These results show promise for gaze-based interaction and suggest that these interactive engagements can enhance the learning experience, although not reduce mind-wandering
Design and Characterization of Peptide-Based Nanoparticles for GRP78-Dependent pDNA Transfection of Prostate Cancer Cells
No full textThis thesis project is based on the synthetic fluorescein-labeled amphiphilic peptides that specifically target csGRP78. The amphiphilic polycationic peptide sequences form stable helical-coiled structures that self-assemble into higher-ordered peptide nanostructures. It also describes the use of condensed nanoparticles for efficient gene delivery and anti-cancer activity in GRP78-overexpressing prostate cancer (DU145) cells. The peptide-based nanoparticle formulation was composed of a GRP78-targeting peptide sequence (W) and the nona-arginine (R9) cell-penetrating domain (W-R9), that effectively facilitated the capture and release of pDNA, while also conferring serum stability, as demonstrated by agarose gel electrophoresis. Confocal imaging of the FITC-labeled peptide:pDNA formulation indicated efficient cell uptake and intracellular localization in entrapped endosomes in the DU145 cells. Optimization studies with the Green Fluorescent Protein reporter revealed spermidine and chloroquine as effective additives for enhancing peptide-based transfection efficiency. Moreover, blocking with a primary GRP78 antibody confirmed a GRP78-dependent mechanism for cell entry, leading to pGFP expression
Characterization and classification of Fiber-like Structures in Cells and Tissues
No full textCells are the fundamental building blocks of life; they play foundational roles in biological processes. Cellular organelles such as mitochondria function to send biochemical signals to allow for energy production. Studies have focused on mitochondria as a unit, but the link between the morphology of the organelles and the effects on cellular processes needs to be better understood. Tissues comprise cells and noncellular components, such as the Extracellular matrix (ECM). The ECM is crucial in scaffolding tissue through proteins such as collagen and elastin. Microscopy images and gray-level texture analysis were used to find differences in mitochondrial structures and fibrillar collagen were used to study mitochondria and fibrillar collagen structures in the context of aging, providing fundamental information for future use in applications such as cellular responses to microenvironmental changes and drug discovery, the imaging modalities used were Super-resolution confocal microscopy and Second Harmonic Generation (SHG) imaging
Improved IMU/GNSS EKF fusion using Machine Learning
No full textInertial Measurement Unit (IMU) sensors are the essential components in navigation systems. Micro-Electro-Mechanical System (MEMS) sensors are a less affordable and lightweight IMU. MEMS suffer from larger stochastic errors that accumulate over time, eventually causing drifts in navigation. The integration of Global Navigation Satellite System (GNSS) measurements with IMU data solves the issue of navigation drift and provides accurate navigation. However, this integration falls short in during GNSS outage scenarios, primarily due to persistent IMU errors. This thesis introduces an effective strategy to reduce navigation drifts by rectifying IMU errors using Machine Learning (ML) algorithms such as Light Gradient Boosting Machine (LightGBM) and Categorical Boosting (CatBoost). In contrast to many other methodologies reliant on high-end and expensive IMUs for denoising low-cost MEMS IMUs, this thesis proposes Inverse Kinematics (IK). The proposed method undergoes testing in Loosely coupled and Tightly coupled schemes under varying GNSS outage durations
Shortest Path Problems in Weighted Regions
No full textFinding a shortest path is one of the most studied problems in computational geometry. There are many variants of the problem. One such variant is the Weighted Region Problem (WRP), i.e., finding a shortest path when the underlying domain is weighted. The difficulty of finding exact weighted shortest paths motivates the study of approximation algorithms for the problem. One alternative that is often encountered is to discretize the continuous space by considering a weighted mesh. Then, optimal shortest paths in this subdivision are approximated. We present two methods that discretize the space based on the placement of Steiner points in the cells of an equilateral-triangle tessellation. Using such a discretization, we can use algorithms for shortest paths in graphs for computing a shortest path in the geometric graph obtained. This will lead us to two approximation algorithms for solving the WRP. In addition, we study how well a weighted mesh approximates the space, with respect to shortest paths. We consider a shortest path~ from~ to in the continuous 2-dimensional space, a shortest vertex path~ (or any-angle path), which is a shortest path where the vertices of the path are vertices of the mesh, and a shortest grid path , which is a shortest path in a graph associated to the weighted mesh. We provide upper and lower bounds on the ratios , , in equilateral-triangle, square and regular-hexagon meshes. Finally, we explore an application of shortest paths applied to a robotics problem. We study the problem of determining minimum-length coordinated motions for two axis-aligned square robots translating in an obstacle-free plane: Given feasible start and goal configurations, find a continuous motion for the two squares from start to goal, comprising only robot-robot collision-free configurations, such that the Euclidean distance traveled by the two squares is minimal among all possible such motions
Sociological Perspectives of Anti-trafficking Organizations: A Case Study of The Salvation Army
No full textThe Salvation Army is a unique case study as it is an international faith-based organization operating in 134 countries. The overarching research for this project is: How does a faith-based organization such as The Salvation Army name and frame the question of human trafficking? I argue that the re-framing of The Palermo Protocol within The Salvation Army highlights the intersection of international policy and religious beliefs in addressing the issue of human trafficking. By expanding the Traditional 4P Framework, The Salvation Army can address the complexities of human trafficking through a lens of moral obligation and faith-based principles. This re-framing is only a part of anti-trafficking efforts but challenges the dichotomy between international policy and religious perspectives on human trafficking