39339 research outputs found
Sort by
Mechanistic Studies of Non-Enzymatic (Per)Oxidation Pathways of Thiols, Sulfenic Acids and Lipids under Biomimetic Conditions
No full textTo facilitate the metabolic processes needed to sustain life, aerobic lifeforms have evolved to utilize atmospheric oxygen (O₂) as the terminal electron acceptor to generate essential energy. While aerobic metabolism is highly efficient compared to its anaerobic counterpart, one of its drawbacks is the potential for the formation of so-called "reactive oxygen species", which engage in a series of chemical processes that lead to the formation of non-enzymatic metabolites, of which some are harmful to cellular proliferation and may act as molecular signatures for pathology. Among these "reactive oxygen species" are organic (hydro)peroxides (ROOHs) derived from lipids and other biomolecules.
Thiols, in the form of the amino acid cysteine and its derivative, glutathione, are key reductants that eucaryotic cells have evolved to utilize to prevent the accumulation of ROOHs, among many other functions. Despite being central to many aspects of biology, the mechanisms of thiol reactivity have been challenging to characterize, as electrophilic reaction intermediates, like sulfenic acids (RSOH) and sulfenyl chlorides (RSCl), are too reactive to be observed directly. To be able to observe these intermediates in aqueous buffer and to develop a better understanding of the mechanistic underpinnings of thiol oxidation, a fluorinated triptycene thiol was synthesized and its reactivity with H₂O₂, ¹O₂ and HOCl was investigated.
Ferroptosis, an iron-dependent form of cellular death, is strongly associated with the accumulation of lipid hydroperoxides (LOOHs) and the chemical processes that occur as a result. Numerous studies support an intricate relationship between a cell's lipidome and its sensitivity to ferroptosis, however, the molecular basis for this relationship remains poorly understood. To determine whether the ability to initiate and promote iron-dependent LPO is dependent on the structure and reactivity of a ROOH, different ROOHs were individually incorporated into eggPC liposomes and their ability to initiate LPO through addition of Fe(II) was investigated using STY-BODIPY as a fluorescent reporter of autoxidation (the spontaneous reaction of a compound with atmospheric oxygen). Results obtained upon screening various iron chelators in place of EDTA strongly support iron-phospholipid association, ligand displacement on Fe(II) by hydroperoxyl and subsequent inner-sphere reductive heterolysis of LOOHs as the primary chemical steps in LPO initiation. The rate of STY-BODIPY oxidation (and hence, LPO) depends on the position of the hydroperoxyl group (-OOH), where the rate increases as the -OOH is further down the alkyl chains of fatty acids. The obtained results corroborate a theory where physical (H-bonding between membrane components and orientation of the peroxyl group of LOOHs and chemical (association of low molecular weight iron species with the membrane and their ability to reduce LOOHs at the interface, competing LOOH decomposition pathways) effects dictate a lipid bilayer's susceptibility to peroxidation
Quantum Effects in Strong Field Physics
No full textWhen matter is exposed to strong fields, electrons are ionized and a number of processes such as high harmonic generation (HHG), above threshold ionization (ATI), non-sequential ionization happen. These processes are usually investigated semiclassically, i.e. matter is treated by quantum mechanics, and radiation is treated classically. In particular HHG which is the generation of ultrashort coherent extreme ultraviolet (XUV) pulses, has enabled a wealth of novel ultrashort spectroscopic methods in atomic, molecular, and condensed matter physics. Further, ionization has applications in micro-machining. Strong field physics continues to have numerous open questions. As an illustration, the mechanisms underlying dominant HHG pathways in solids remain insufficiently understood. It is not clear which mechanism dominates HHG in various materials.
Possible pathways are the interband and intraband contributions. Furthermore, many of these processes in strong field physics have been described by single active electron (SAE) approximation. This approach relies on the assumption that the interaction between a single electron and the intense laser field dominates all other interactions. However, this is clearly not the case especially in solids where interactions with the lattice and other electrons can be important. In addition, quantum optical aspects of strong field physics have generated increasing interest, yet they remain partially explored. This chapter involves treating electromagnetic fields using the formalism of second quantization. Quantum optical properties, such as squeezing, entanglement and the negativity of the Wigner function, are of fundamental importance for the field of quantum information and quantum computation. The central theme of my thesis is the development of quantum optical and statistical theoretical frameworks for describing intense laser field processes such as ionization and
HHG in atomic and molecular gases and in solids. The second chapter focuses on gaining a detailed understanding of the mathematical steps involved in the Keldysh theory of ionization to establish a solid theoretical foundation. Through this analysis we identified an additional factor of two compared to Keldysh’s original derivation of atomic ionization rates. The third chapter addresses the inadequately understood mechanisms that govern dominant high harmonic generation (HHG) pathways in solids. One approach to clarify these mechanisms involves introducing real (resonant) and virtual processes. We developed the strong field adiabatic following (SFAF) formalism which is based on Dyson expansion using the von-Neumann equation of density matrix. Using SFAF a diagnostic method is obtained to separate virtual and resonant channels. Through this separation, and by comparing with experimental results, we identified the need to incorporate many-body
effects. The fourth chapter explores the fact that a solid is a complex many body system in which the SAE approximation is very crude. The electron interacts with holes, other electrons, collective excitations such as plasmons, and phonons. Our work is based on the idea that all theses effects can be treated as a quantum statistical heat bath of bosonic harmonic oscillators. We apply this to our SFAF formalism. In the ionization case, this work has settled a long standing issue which is the fact that simple phenomenological approaches such as the relaxation time approximation 2 result in nonphysical enhancement of ionization.
In chapter five, we have investigated ways to transfer quantum optical properties on the otherwise classical high harmonic radiation. The goal is to use HHG to scale quantum sources to smaller wavelengths. Quantum properties can be imprinted by perturbing HHG with a quantum field, such as bright squeezed vacuum (BSV). In this chapter, the theory of quantum sideband high harmonic generation (QSHHG) in atoms and solids is
derived to find ways by which to transfer quantum properties from the perturbative BSV to the harmonic sideband. The theoretical framework is a quantum generalization of the semi-classical Lewenstein model of HHG. It gives closed-form solutions for the HHG and QSHHG wavefunctions. Knowing the wavefunction, we can identify the quantum properties of QSHHG. The additional photons absorbed and emitted from the quantum perturbation, here BSV, create entanglement between individual harmonic sidebands and between the harmonic sidebands and the BSV. We show how this entanglement can be used to create a variety of non-classical states commonly used in quantum information science, such as high-purity single-photon states, Schrödinger cat states, and photon-added squeezed vacuum states. Some of these non-Gaussian states with negative Wigner functions have been shown to provide a quantum computational advantage over their Gaussian counterparts. Additionally, they play a significant role in quantum metrology, enhancing precision measurements beyond classical limits. This chapter has opened a path to quantum engineering of HHG
Transformer and Graph Neural Networks Based Vulnerability Detection in Binary Code
No full textThe growth of the Internet of Things (IoT) has heightened the need for effective binary-level N-day vulnerability detection, particularly when source code is unavailable. The binaries deployed on IoT devices are often compiled under varying conditions, such as different compilers and optimization levels, and, in cross-platform settings, across multiple CPU architectures. This high degree of variability means that even functionally identical code can have vastly different binary representations, posing significant challenges to conventional code analysis techniques (e.g., simple syntactic or signature-based analysis methods). Existing ML-based methods typically rely either on structural graph representations binary code or on instruction sequences of the code, each capturing only a partial view of a binary's semantics.
This thesis introduces a hybrid analysis framework that integrates both structural and semantic perspectives using the Code Property Graph with Natural Code Sequence (CPGNCS), with the goal of improving binary vulnerability detection across diverse compilation settings. This graph-based representation encodes syntax, control flow, and data dependencies, while preserving the natural order of instructions. To enhance semantic understanding within individual nodes in the CPG-NCS graph, we incorporate CodeBERT, a transformer-based language model pretrained on code. These enriched node embeddings are processed by a Gated Graph Neural Network (GGNN) trained in a Siamese architecture to identify functional similarity across binary functions.
Our method achieves significant improvements in vulnerability detection performance under diverse compilation settings. Notably, it yields an average increase of 5% in F1-score and up to 8% in AUC compared with state-of-the-art solutions. We also test our approach on a variety of obfuscated code built using different obfuscators. Additionally, we introduce a formulation for estimating the number of GGNN layers based on graph feature metrics and validate its effectiveness through extensive experiments. The practical utility of our approach is further demonstrated by successfully identifying known vulnerabilities in real-world IoT firmware images (e.g. IoT operating system)
Multimodal Emotion Recognition Using Physiological Signals
No full textAffective computing aims to develop systems capable of recognizing and interpreting human emotions, yet existing multimodal datasets frequently suffer from limitations such as poor signal quality, high inter-subject variability, and inconsistent evaluation protocols. To address these gaps, this thesis develops and validates a comprehensive framework for multimodal emotion recognition using physiological signals - Electrocardiogram (ECG), Electrodermal Activity (EDA), and Respiration (RSP) - augmented with speech-based representations. The goal was to establish standardized preprocessing workflows, rigorous signal quality assessment (SQA), and reproducible baseline experiments to support the development and technical validation of a large-scale physiological dataset. This framework was applied to a dataset collected from 99 participants, containing synchronized physiological recordings, speech responses, and self-reported emotional annotations during exposure to validated video stimuli.
To ensure data integrity, a rigorous SQA and artifact-removal pipeline was applied across modalities, integrating established ECG and respiration metrics with newly designed EDA-specific indicators. Using this refined dataset, multiple emotion-classification experiments were conducted under a strict subject-independent evaluation protocol, comparing fixed 30-second windows with emotion-triggered temporal segments. Across all tasks - binary arousal, binary valence, and multiclass emotion recognition - trigger-based segments consistently produced clearer and more discriminative physiological patterns. Random Forest achieved the strongest overall performance, including 78.8% multiclass accuracy using physiological features alone.
To explore multimodal enhancement, speech embeddings were fused with handcrafted physiological features. This early-fusion approach led to substantial improvements across all tasks, most notably increasing multiclass accuracy from 78.8% to 97% when using trigger-based segments. These findings demonstrate that speech provides complementary affective information that enhances physiological representations. A subject-wise evaluation was also conducted to examine emotion separability across individuals and to identify video-specific misclassification patterns that reveal how different stimuli elicit varying physiological responses. Overall, this thesis delivers a validated multimodal dataset, reproducible processing pipelines, and strong baseline benchmarks that provide a solid foundation for future research in physiological and multimodal emotion recognition
Neural Features of Obesity and Intervention: A Food-Cue fMRI Study
No full textObesity is a global epidemic associated with alterations in reward processing and inhibitory control. Pharmacological interventions, such as Contrave® (naltrexone/bupropion), combined with dietary modification, may influence these neural processes; however, the short-term neural effects of such interventions remain unknown. This is a double-blind, placebo-controlled, and randomized phase four clinical trial investigating neural responses to food cues before and after a 4-week Contrave® and diet intervention (total N=15 at baseline and N=11 pre/post). At baseline, we identified activity in several regions (notably the insula) across different food-viewing conditions. Baseline brain responses to food cues were also associated with in-scanner food cue ratings in the Rolandic operculum, as well as reward motivation traits in the dorsolateral prefrontal cortex, and more. The 4-week intervention induced changes in brain activity in the hippocampus and precuneus, among other regions. Changes in brain activity were also associated with reward motivation traits in several regions, including the insula and frontal cortex. This study contributes evidence regarding the neural correlates of obesity and the potential impact of a short-term drug/diet intervention. To our knowledge, it is the only study to combine fMRI with a pharmacological and dietary weight-loss intervention in men and women with obesity. Overall, we hope that this research will improve obesity-related patient care and the clinical use of Contrave®
Structural Agent-Based Approach for 3D Geological Surface Modelling
No full textGeological surface modelling is the process of creating 3D digital representations of the Earth's subsurface structures and surfaces based on geological, geophysical, and geospatial data. It's commonly used in geology, mining, petroleum exploration, hydrology, and geotechnical engineering. Traditional interpolation methods often rely on assumptions like spatial stationarity and high data availability, which are not always the case in earth sciences applications. These methods fail to capture local variation in geologically complex and data sparse areas, as they tend to be biased towards global means. This research explores how agent-based modelling (ABM) integrated with statistical and domain-specific knowledge can be used to improve geospatial estimation when data is sparse and incomplete, especially in geological surface modelling. In this study, structural agents are introduced as independent entities that interact with each other using rules from the classic BOID flocking algorithm: cohesion, separation and alignment. Each agent is defined by its spatial position (X,Y,Z), a normal (N₁,N₂,N₃) and a velocity (V₁,V₂,V₃) orientations, and updates its behaviour over time through local communication. A two-phase approach was used: Phase I tested if agents could self-organize, and Phase II evaluated how well agents could propagate information from real data points. The long-term goal of the study is to establish viability of a method for reconstructing surfaces of complex geological features such as faults, poly-deformed folds and regional fault networks. In the short term, this work focuses on determining what data configurations and agent behaviours support the simplest reconstruction scenarios, before addressing more complex geological cases
A stamp of approval is not a seal of ethics: an editorial on justice and respect in trial design
No full textDépendance et Mutations du processus de production et de circulation du savoir à l'ère numérique: le cas de la revue savante de l'Université Félix Houphouët Boigny et de l'Université du Ghana
No full textLa transition des revues savantes vers le numérique nécessite des compétences individuelles et organisationnelles. Tandis que les pays industrialisés avancent dans ce processus en numérisant leurs livres et revues, les changements en Afrique restent très peu méconnus. Cette recherche fournit des informations descriptives sur les revues savantes en français et en anglais et de l'usage de ces langues dans les thèses de doctorat de 2016 à 2020 de l'UFHBCI et de l'UG pour comprendre les facteurs limitants et encourageants de la production et de la circulation des savoirs. Notre étude a pour objectif de comprendre : comment l'évolution de nouvelles pratiques technologiques, linguistiques et économiques permet-elle d'appréhender les facteurs qui limitent ou qui encouragent une production locale du savoir? Une approche mixte (qualitative et quantitative) a permis d'examiner les trois dimensions et des stratégies mises en place pour une adaptation aux mutations du processus de production et de circulation du savoir à l'ère numérique. L'étude constate la transition des modèles économiques des revues vers le libre accès. En attendant des recherches plus étendues avec échantillonnage plus large, l'analyse basée sur les données que nous avons collectées révèle que l'enseignement supérieur dans ces pays reste influencé par les anciennes puissances coloniales, notamment à travers le financement, l'usage des langues et l'adoption de technologies étrangères. Cette dépendance se traduit par une forte utilisation majoritaire de ressources académiques externes, favorisant une production scientifique plus tournée vers le savoir global que local. Enfin, le sous-financement apparaît comme le principal frein au développement de la recherche, limitant l'autonomie institutionnelle et la production de contenus adaptés aux réalités locales. Bien restreintes, les données recueillies dans notre étude donnent un aperçu du financement octroyé par leurs dirigeants respectifs
Dynamic Response and Design of Steel Structures Subjected to Pedestrian-Induced Loads
No full textThe present thesis investigates the dynamic performance of steel members under the effect of pedestrian induced dynamic forces. Towards this objective, the study adopts two main methodologies: (a) controlling the member natural frequencies to mitigate possible resonance phenomena under human-induced dynamic forces and (b) controlling acceleration levels induced in members to ensure they fall below acceptable perception levels.
While present design Canadian and American structural steel design provisions do not offer explicit natural frequency requirements, they limit member slenderness to 200 for compression members and 300 for tension members. These limits partly control the natural frequency but omit the influence of axial load level, and neither fully capture the effect to cross-sectional asymmetry, as may be the case in angle members, nor fully account for member end-connection details. Within this context, the present research presents four contributions towards advancing the state of the art related to the analysis and design of steel members under human activities.
The first contribution examines the natural vibration response of axially loaded members with wide-flange sections. A parametric study is conducted using closed-form analytical solutions and shell finite element modelling. The study characterizes the effect of axial loads on the member natural frequencies and proposes slenderness limit thresholds based on target natural frequencies that capture the axial load level. In this respect, this part of the study develops an improved serviceability criterion than simply satisfying member slenderness requirements in the present standards.
The second contribution extends the investigation to members with angle cross-sections, which are prone to torsional–flexural coupling owing to cross-sectional asymmetry, an aspect not addressed in the first contribution. A closed-form solution is developed to determine the torsional flexural natural frequency of pin-ended members with equal leg angles. The analytical model is complemented with shell finite element modelling to tackle more general cases involving unequal leg angles and members with gusset plate end connections. A subsequent parametric investigates the effect of axial force, cross-sectional geometry, member span, and boundary conditions on the natural frequency of angle members. The findings show that satisfying present slenderness criteria does not guarantee consistent natural frequencies across members with angle cross-sections and flags the need to develop more elaborate criteria to control excessive vibrations.
The third contribution formulates a general thin-walled beam finite element formulation for the natural vibration analysis of steel members. The solution captures torsional-flexural coupling induced by cross-sectional asymmetry, and the effects of axial loading, warping, and rotary inertia. The formulation is equipped with a feature that enables the seamless modelling of the boundary conditions of gusset-plate ended connections, in which the member is considered fixed about the axis normal to the gusset but nearly pinned about the gusset axis, both axes being non-principal. The findings reveal that gusset plate end connections significantly elevate the natural frequency of the member, when compared to pinned-ended members. A simplified energy-based solution is also developed to estimate the natural frequency of members with gusset-end connections, and a dimensionless design chart is provided to streamline the calculation procedure in a design environment.
While the previous three contributions centre around quantifying the natural frequency of steel members, the final contribution transitions to the full transient response evaluation under the time-
dependent loading is induced by human walking. Towards this objective, a general purpose thin-walled beam finite element formulation is developed for the fully dynamic analysis of thin-walled beams. In addition to capturing the effects of coupling induced by cross-sectional asymmetry, axial loading, warping, and rotary inertia, the formulation incorporates the effect of damping and develops the energy equivalent force vector that characterizes the temporal and spatial distributions of forces induced by human walking. The discretized equations of motion are then solved using the Newmark time integration scheme, and the predictions of the model are validated against benchmark solutions. A parametric study explores the effects of axial force magnitude, damping model, damping ratio, boundary conditions, and characteristics of the walking function on the member acceleration response. The results are benchmarked against established human comfort thresholds. The model thus provides a basis to assess the dynamic performance of steel members against established acceleration thresholds.
In summary, the present thesis advances methods of dynamic analysis of steel members. By integrating analytical solutions, energy-based approximate methods, and finite element modelling techniques, the study offers a possible framework for the analysis and design of steel members under the effect of dynamic loads induced by pedestrian-induced loads
Synergizing Experimental and Computational Methods for Gas Separation Membrane Characterization
No full textThe field of membrane science has experienced remarkable advancements in gas separation technologies recently. A key challenge in enhancing membrane performance is exceeding the "upper bound" limit, which requires simultaneously high permeability and selectivity. To meet this goal, reliable and swift membrane characterization techniques are critical in the process of developing new materials for gas separation membranes. The time-lag method is the most widely used technique for determining the gas transport coefficients in gas separation membranes, which allows evaluating membrane permeability (P), diffusivity (D), and solubility (S) coefficients in a single permeation experiment. However, the conventional time-lag method is limited by the accuracy of the monitoring system, and when applied to more complex cases or more advanced materials. The objective of this thesis is to explore more efficient and reliable membrane characterization techniques to overcome these challenges.
In this project, we developed and employed a novel constant-volume (CV) system that enables simultaneous monitoring of dynamic upstream pressure decay and the downstream pressure rise. This innovative CV system allowed us to leverage both upstream pressure decay and downstream pressure rise to propose new membrane characterization methods for a more robust determination of P, D, and S. The resolution and the accuracy of the system were significantly improved by splitting the upstream into the working volume and the reference volume. A new experiment protocol was also presented to minimize the impact of the adiabatic expansion during the initiation of the test. Thanks to the advantage of the system, we proposed and validated a series of new methods using a rubbery membrane, namely polydimethylsiloxane (PDMS). The transport properties obtained through these new methods yielded transport coefficients that were very close to those determined using the conventional time-lag method.
To better calibrate the system, this work also studied some potential artifacts in the system. The effect of using a porous support disc on estimating the membrane's effective transport parameters was explored by numerical simulations. The results show that relative diffusivity and relative permeability are strictly a function of the porosity of the porous disc and the ratio of the pore diameter to the membrane thickness, which can be used to correct the impact of the porous plate and recover the intrinsic membrane properties.
Based on the improved experimental system and analytical protocol, we investigated the influence of the applied pressure on membrane transport properties due to the phenomenon known as the membrane compaction effect. To systematically evaluate this effect, we first measured the thickness of the compressed membranes under different loads in a high-precision testing machine. Next, we utilized our novel CV system to characterize the membrane transport properties under different pressure conditions. The experimental results revealed an apparent reduction in the transport properties of the PDMS membrane as it underwent compression due to gas pressure differentials.
To determine the most probable values of the transport, a data reconciliation technique was employed, considering both methods and measurement deviation. This approach minimizes an objective function that integrates all sources of variation. A gradient descent optimization algorithm was then applied to converge on the best estimates of P, D, and S. The reconciled results improve the consistency and precision in estimating transport properties compared to the original results.
In parallel with our experimental work, we conducted numerical simulations, including the finite difference (FD) method, electrical analogy networks, and the Monte Carlo (MC) technique, to complement and enhance our findings. These simulations generated upstream and downstream pressure-time profiles corresponding to membrane behavior, providing valuable insight into the actual experimental data. To examine the effect of a porous support disc on the membrane's effective transport parameters under ideal conditions, we solved Fick's second law of diffusion using the finite difference method. The results clearly showed that the membrane thickness, the pore size, and the disc porosity can significantly influence the estimation of the diffusivity and permeability of the membrane when using the time-lag method. These findings not only help calibrate our new CV system more accurately but also offer theoretical support for characterizing more complex materials, such as glassy polymers and mixed-matrix membranes