Recherche uO, le dépôt numérique de l'Université d'Ottawa, réunit le matériel de recherche et d'enseignement créé par notre communauté universitaire et nos partenaires. Le savoir de l'Université est ainsi disponible à long terme et en accès libre, ce qui lui procure de la visibilité et facilite sa diffusion.

Comment soumettre votre thèse
Comment déposer votre recherche

Communautés dans DSpace

Sélectionner une communauté pour parcourir les collections sous-jacentes.

Nouveaux dépôts

  • Vignette d'image
    Type d'Item : Item ,
    Cover-free Families on Graphs and Hypergraphs
    (Université d'Ottawa | University of Ottawa, 2026-09-03) Parida, Prangya; Moura, Lucia
    A family of subsets of a $t$-set is a \emph{$d$-cover-free family} or $d$-CFF if no subset in the family is contained in the union of any $d$ other subsets. Let $t(d, n)$ denote the minimum $t$ for which there exists a $d$-CFF on a $t$-set with $n$ subsets. Since a $1$-CFF is the same as a Sperner family, using Sperner's theorem, we get $t(1, n) \sim \log_{2}(n)$ as $n$ grows. Erd\H{o}s, Frankl, and Füredi (JCTA 1982) proved that $3.106\log_{2}(n) < t(2,n) < 5.512\log_{2}(n)$. This thesis focuses on generalizing $d$-CFF using a graph and hypergraph where vertices correspond to subsets in the set system. The main contributions of this thesis are in three main topics. First, we focus on generalizing $1$-CFF and $2$-CFF using a graph $G = ([1, n], E)$ where a subset in the family corresponds to a vertex of $G$. A $G$-Sperner$(t, n)$ is a family of subsets of a $t$-set such that each edge of $G$ specifies a pair of subsets that must not be contained in each other, while a $G$-CFF$(t, n)$ is a family of subsets of a $t$-set such that it is $G$-Sperner and the union of each pair of subsets corresponding to an edge of $G$ does not contain any other subset in the family. Let $t_s(G)$ and $t(G)$ denote the minimum $t$ for which there exist a $G$-Sperner$(t, n)$ and a $G$-CFF$(t, n)$, respectively. In this way, $t_s(K_n) = t(1, n)$ and $t(K_n) = t(2, n)$. Firstly, we prove $t_s(G) = t(1, \chi(G))$ for any simple graph $G$ with no isolated vertices, and provide various upper and lower bounds for $t(G)$. The \emph{trivial bound}, $t(1, n) \leq t(G) \leq t(2, n)$ holds for any simple graph $G$ with no isolated vertex, with the lower bound tight for an infinite family of star graphs and the upper bound tight for complete graphs. We study when these bounds can be improved and give better constructive upper bounds for families of graphs such as stars, paths, cycles, wheels, and windmill graphs. In particular, a construction based on mixed-radix Gray codes yields $\log_{2}(n) \leq t(P_n) \leq t(C_n) \leq 1.893\log_{2}(n) + \O(1)$ where $P_n$ and $C_n$ are paths and cycles with $n$ vertices. Second, we study a generalization of $d$-CFF for $d \geq 2$ using a hypergraph $H$ of rank $d$ and $n$ vertices. An $H$-CFF$(t, n)$ is a family of $n$ subsets of a $t$-set such that the union of the subsets corresponding to the vertices in a subset of a hyperedge does not contain any other subset in the family. Let $t(H)$ denote the minimum $t$ for which there exists an $H$-CFF$(t, n)$. In this way, $t(H) = t(d, n)$ if $H$ is a complete $d$-uniform hypergraph on $n$ vertices, but can be much smaller for hypergraphs in general. We study the connection between $H$-CFFs and other generalized set systems used in group testing, such as $H$-separable set systems, investigate their characterizations through directed hypergraph homomorphisms, and explore bounds on $t(H)$ using tools from hypergraph theory. Furthermore, we study CFFs on graph and hypergraph products. We provide an upper bound for CFFs on the Cartesian product of (hyper)graphs and investigate families of graphs that attain the upper bound, as well as families for which the upper bound is close to the trivial lower bound. Moreover, we generalize some well-known constructions of $d$-CFFs through the lens of CFFs on the strong product of (hyper)graphs. Finally, we focus on the classical mixed-radix \emph{reflected} and \emph{modular} Gray codes by providing their recursive constructions, which are necessary for proving a result concerning cover-free families on paths and cycles. Their respective loopless algorithms can be found as Algorithm H and Exercise 77 in Section 7.2.1.1 of The Art of Computer Programming Vol. 4A by Knuth. Furthermore, we study their generation through the lens of their change sequences. We observe that both of these Gray codes, built on a mixed-radix base, are guided by the same change sequence, called the \emph{ruler sequence}. We show how the generation of the reflected and modular Gray codes can be fully parallelized, generating each new codeword in constant time. We also show that modular Gray codes can be generated using a greedy cyclic increment approach. Furthermore, we present a new family of modular Gray codes starting from any tuple $w$ that can be constructed using the same greedy cyclic increment approach to generate the next word that is lexicographically greater than or equal to $w$. We show that although this order is not a suffix of the original modular Gray code, its change sequence is a suffix of the latter.
  • Vignette d'image
    Type d'Item : Item ,
    Dangerous Technologies and the Loss of Privacy Rights: Analyzing Canadian Lawful Access Legislation through Social Problem Theory
    (Université d'Ottawa | University of Ottawa, 2026-09-03) Nellis, Brent; Steeves, Valerie
    This project analyzes efforts to “modernize” Canadian lawful access legislation, focusing primarily on legislation proposed by the Conservative government between 2009 and 2014. Though privacy is a democratic right integral to citizens’ political participation, research shows that governments often mobilize security threats to justify the expansion of surveillance measures that infringe upon privacy rights, citing the need to “balance” national security and civil rights. The purpose of this project is to understand how security and privacy were socially constructed in relation to lawful access legislation. Using social problem theory, which studies how social problems are socially constructed through the rhetorical claims-making activities of relevant parties, this project examines newspaper articles, parliamentary records, and government documents to understand how political parties, lawyers, privacy advocates, victims’ advocates, and other claims-makers defined the security and privacy issues that surrounded lawful access legislation. Similarly to broader trends, the endangerment of Canadian society by cybercrime and cyberbullying and the loss of privacy rights were prominent issues in the rhetoric, as was the need to balance security and privacy. However, the analysis shows that proponents of lawful access legislation took a nuanced approach to the balance between security and privacy by claiming that the legislation would respect and protect privacy rights. This rhetorical claim subverted the common framing of security and privacy as mutually exclusive benefits in a zero-sum game. This suggests that defending privacy rights on the grounds that they are equal to security rights would be an ineffective strategy, as the state can simply deny claims that its policies would infringe upon privacy rights.
  • Vignette d'image
    Type d'Item : Item ,
    DPP4 Links Metabolic Disease and Atherosclerosis: Regulation of the Inflammatory Plaque Microenvironment
    (Université d'Ottawa | University of Ottawa, 2026-09-03) Abdalbari, Ali; Mulvihill, Erin
  • Vignette d'image
    Type d'Item : Item ,
    Characterizing the Role of RNA-Binding Protein Human Antigen R in the Neuromuscular Junction in Duchenne Muscular Dystrophy
    (Université d'Ottawa | University of Ottawa, 2026-09-03) Bhaura, Gurleen; Jasmin, Bernard
    Duchenne muscular dystrophy (DMD) is a progressive X-linked disorder caused by mutations in the dystrophin gene, leading to severe skeletal muscle fibre degeneration and neuromuscular junction (NMJ) abnormalities, including fragmentation and disorganization of the post-synaptic apparatus. Human Antigen R (HuR) is an RNA-binding protein involved in post-transcriptional regulation of transcripts associated with muscle differentiation and endplate function. However, its role in maintaining the structure of the post-synaptic apparatus at the NMJ remains unclear. This project investigated HuR's contribution to dystrophic endplate organization in the mdx mouse model of DMD. First, we confirmed significant post-synaptic disorganization in the mdx muscle relative to wild-type (WT) controls. Additionally, HuR protein expression was elevated (~3-fold) across multiple fast- and slow-twitch mdx muscles from 4-12 weeks, with smaller, non-significant increases observed at 24+ weeks. To further characterize HuR localization, subcellular fractionation was performed in the tibialis anterior (TA) muscle at 11 weeks, a time point representative of the increased HuR expression observed across the multiple muscles and ages we examined. Subcellular fractionation demonstrated significantly increased nuclear HuR protein levels in mdx muscle compared to WT controls (p<0.001). Consistent with these findings, immunofluorescence analyses also revealed a strong trend towards increased nuclear HuR abundance in mdx muscle fibres relative to WT controls in the TA when all ages were combined (p=0.054). Given these findings, we knocked down HuR expression in vivo using recombinant adeno-associated virus-mediated short hairpin RNA delivery through intramuscular injections in the TA muscles of mdx mice. HuR protein expression was reduced by approximately 60% relative to shLuc-treated controls, restoring HuR levels to those observed in WT muscle (p<0.01). This led to a partial improvement in post-synaptic organization, as evidenced by increased acetylcholine receptor compactness compared with shLuc-treated mdx muscle (p<0.05) and untreated mdx muscle (p<0.0001). Finally, the pan-histone deacetylase inhibitor Vorinostat was evaluated as a pharmacological approach to assess changes in HuR expression, localization, and potential effects on post-synaptic morphology. However, Vorinostat did not significantly increase acetylated tubulin levels, suggesting that adequate HDAC inhibition may not have been achieved, which limited our ability to assess its effects on HuR expression and post-synaptic organization. Overall, HuR protein expression is dysregulated in dystrophic muscle across multiple ages and muscle types, while reducing HuR expression partially improved post-synaptic membrane organization, supporting a role for HuR in the molecular mechanisms underlying endplate abnormalities in DMD.
  • Vignette d'image
    Type d'Item : Item ,
    TRACE: A Multimodal Multi-Agent System for Transparent Chest X-Ray Assistance
    (Université d'Ottawa / University of Ottawa, 2026-09-02) Wang, Zhaoyu; El Saddik, Abdulmotaleb
    Chest X-ray (CXR) AI has shown strong progress in abnormality detection, report generation, and multimodal medical reasoning, yet many existing systems remain limited as user-facing assistive tools. In particular, report-centered or text-only systems often provide limited support for evidence inspection, follow-up interaction, and cross-study continuity. This thesis presents TRACE (Transparent Radiology Assistance for Chest X-ray Explanation), a multimodal multi-agent system for transparent CXR assistance. TRACE integrates detector-supported image analysis, multimodal report drafting, report refinement, transparency-oriented interaction, consultation support, and archive-assisted comparison within a unified workflow. TRACE was evaluated from four complementary perspectives: objective report quality, transparency-supported consultation preparation, archive-assisted cross-study comparison, and subjective usability and workload. Objective report quality was assessed on a 2,000-image frontal CXR benchmark constructed from MIMIC-CXR and the Indiana University Chest X-rays dataset, using external comparisons against LLaVA-Med and the MedRAX report-generation tool as well as an internal ablation of the TRACE reporting pipeline. TRACE achieved the best overall profile across most reported semantic and label-oriented metrics, indicating that detector-supported finding integration and report refinement can improve report quality within a broader assistive workflow. Two controlled user studies and a questionnaire-based human-factors assessment were also conducted with 48 participants. The transparency-oriented interface significantly improved users' ability to identify verification-worthy report conclusions, justify why they warranted clinician verification, and formulate more specific verification-oriented consultation questions. Archive support significantly improved report-level difference identification and reduced cross-study comparison time. In addition, the full TRACE interface was associated with higher perceived usability and lower subjective workload than the text-only interface, with the largest workload reductions observed in mental demand and frustration. Overall, this thesis shows that, within the controlled benchmark and user-study settings examined here, CXR AI can provide measurable assistive value when it is designed and evaluated not only for report generation, but also for evidence inspection, verification-oriented interaction, and continuity across encounters within a unified assistive workflow. Clinical usefulness remains to be established through further validation with patients and clinicians.