143174 research outputs found
Sort by
Characterizing pyroelectric detectors for quantitative synchrotron radiation measurements
No full textThis paper describes the fabrication and quantitative characterization of pyroelectric detectors for the measurement of the intensity of synchrotron radiation in the challenging hard X-ray region (5–20 keV). The measurement of this radiation requires robust detectors with high signal-to-noise ratios. Our study examines the
response and noise characteristics in this spectral region of pyroelectric detectors made with three contrasting
ferroelectric materials: LiTaO3 (LT), LiNbO3 (LN) and triglycine sulphate (TGS). The key parameters of voltage
responsivity(Rv), noise equivalent power (NEP), and detectivity (D*) are analysed across a frequency range of
1 Hz to 100 Hz, with a photon energy of 17 keV. The detector made with TGS emerges exhibited the best
radiometric performance with an Rv of 7.09 × 10 ³ V/W and an NEP 1.75 × 10⁻⁸ W/√Hz at 10 Hz, which is
comparable with conventional X-ray scintillation detectors. In comparison, a detector using LT demonstrated an
Rv of 1.8 × 103 V/W and an NEP of 5.02 × 10− 8 W/√Hz under similar conditions. The LN-based device was ca
3–4 times lower in response than the LT device. All experimental measurements showed excellent agreement
with theoretical predictions, indicating that predictions of potential device design improvements using these
models should be highly reliable. The LT devices showed excellent linearity of response. LT also possesses a much
higher depolarization temperature and is considerably more robust than TGS. These findings open new avenues
for enhancing X-ray detection capabilities, particularly in challenging synchrotron environments where traditional detectors may face limitations
New alluaudite-type Na3.4Co1.3(MoO4)3 for sodium-ion batteries and electrocatalytic hydrogen evolution
No full textThe alluaudite-type compound Na3.4Co1.3(MoO4)3 has been successfully synthesized by a solid-state process route. It crystallizes in the monoclinic system (C2/c), cell parameters (Å,°): a = 12.582(5) Å, b = 13.449(8) Å, c = 7.119(7) Å, β = 112.02(4)°, V = 1116.8(14) Å3, and Z = 4. Its crystal structure consists of octahedral [MO6] (M = Co, Na) and tetrahedral [MoO4] that share corners and/or edges to build the 3D framework. The sample was also characterized by X-ray powder diffraction, which confirmed crystal data, and infrared (FT-IR) and Raman spectroscopies. Its morphology was analyzed using scanning electron microscopy (SEM). The vibrational study confirms the existence of the MoO2-4 functional groups. The title compound was determined to be paramagnetic. In addition, the compound was characterized by cyclic voltammetry (CV), and its electrochemical performance and impedance were analyzed by electrochemical impedance spectroscopy (EIS). The electrochemical reaction mechanism and the limiting factors of Na3.4Co1.3(MoO4)3 as electrode material in Na-ion batteries at room temperature were also discussed. The prepared electrocatalyst showed modest hydrogen evolution reaction (HER) performance with an overpotential of 309 mV required to afford a current density of 10 mA cm−2
Optimising data collection strategies in cyber security to address bias
No full textCyber attacks on digital systems continue to increase in diversity, sophistication, and volume. Consequently, defensive measures must be adaptive, rapid, and automated. By generalising from readily available attack data, machine learning (ML) technologies have proven useful in automating the detection, identification, and characterisation of threats to reduce the adverse impact of attacks. The performance of ML depends on whether its training data contains important patterns it needs to learn or else “garbage in, garbage out.” This thesis seeks to improve the performance of ML models in cyber security by optimising data collection strategies, emphasising eliminating the potential impact of biases.
First, we craft and deploy fully instrumented cloud-based honeypots with a unique class of kernel-level sensors. We have published millions of host and network events with in-the-wild attacks. Second, to optimise honeypot deployments and any other intrusion data collection system, we provide a rigorous method to identify bias using causal graphs. We demonstrate how data collections that target bias improve ML model performance compared to only collections within the same environment. Finally, we developed an adaptive experimental design to reduce the cost of collecting additional data addressing bias. This novel method autonomously alters resource allocation based on the events seen during collection.
Our contributions provide a framework for security experts to better understand and update their datasets. The first contribution provides the data collection tool and initial dataset. The second guides experts to list actionable interventions: feasible changes to explore how an environment generates the data and remove bias that can confuse ML models as they learn. Our final contribution efficiently applies the first automated control trials, comparing the original “control” environment to a new “treated” environment where one listed intervention is applied. By adhering to our process, the updated datasets will encourage ML models to effectively defend digital systems.Open Acces
Adaptive modelling, optimization and control of large-scale systems using machine learning
No full textThis thesis explores the intersection of machine learning (ML), optimization, and control within process systems engineering (PSE), focusing on the real-time optimization (RTO) and control layers to advance industrial autonomy. The primary goal is to enhance adaptability, robustness, and efficiency through ML-driven frameworks.
To address key challenges—plant-model mismatch, computational complexity, and robustness to disturbances—this work develops novel methods across three pillars: data-driven modelling, adaptive systems, and large-scale process control. The research begins with adaptive RTO using Gaussian Processes (GPs) to address model mismatch. This is extended to Adversarially Robust Real-Time Optimization and Control (ARRTOC), which shifts the robustness burden from control to RTO, improving operability under uncertainty. The combined use of GP adaptation and ARRTOC addresses both mismatch and robustness.
In dynamic systems, we conduct a rigorous evaluation of data-driven models using defined metrics to guide model selection and assessment. This study highlights trade-offs across the modelling lifecycle, offering practitioners practical insights.
The thesis further investigates deep learning-based Koopman operator methods to linearize nonlinear dynamics for efficient model-based control. For high-dimensional systems, we introduce ARRO-MPC, an Adversarially Robust Reduced-Order Model Predictive Control framework that uses adversarial training to improve ROM robustness, ensuring efficient yet reliable control.
Overall, the findings show ML can significantly enhance adaptability and robustness in PSE. However, successful deployment hinges on integrating ML with domain expertise. This work bridges ML, control, and optimization to provide a pathway toward more autonomous and resilient industrial systems.Open Acces
High-fidelity modelling of impact breakage in percussive drilling
No full textThe accurate simulation of complex rock failure behaviour in excavation and comminution (e.g., rock crushing and pulverisation) has long been a challenge, particularly in light of the growing demand for subsurface resources and energy. High-fidelity modelling is essential in this field as it enables a deeper understanding of rock failure mechanisms, facilitating the optimisation of drilling operational parameters and tool design. In this research, the combined finite discrete element approach was applied for the first time to rotary percussive drilling simulations, providing a high-fidelity numerical modelling framework in this domain. The integration of a strain-rate effect model and a friction-related pulverisation model enabled the simulation of distinct failure modes in various hard rocks under dynamic point loading. A robust multi-criteria validation approach was proposed and implemented for three types of hard rocks subjected to impacts from hemispherical and ballistic inserts to validate the single-impact model. Compared to conventional single-criterion validation methods based on quasi-static experiments or impact tests, the multi-criteria validation approach demonstrated the high fidelity of the simulations by verifying several key aspects, including bit motion, fragment mass, crater morphology, radial crack length, the sequence of failure modes, and internal crack patterns. Leveraging this high-fidelity model, the energy evolution and dissipation processes during impact-induced rock failure were analysed for the first time, providing new insights for optimising energy utilisation. Additionally, the role of shearing introduced by bit rotation during bit-rock interaction was investigated, uncovering its contribution to the fragmentation process. Building on the validated single-impact model, a multi-impact model was developed to represent the percussion action of a multi-insert bit to study the influence of drilling operational parameters on the drilling rate. This model serves as a valuable tool for optimising operational parameters based on the mechanical properties of different rock types.Open Acces
Mechano-regulation of β2AR signalling pathway of ventricular cardiomyocytes in healthy and failing hearts
No full textMyocardial stretch occurs during haemodynamic overload and may be physiological (e.g., exercise) or pathological (e.g., heart failure). Understanding the effects of stretch on signalling pathways associated with mechanosensitive structures, such as caveolae, is therefore essential. Caveolar microdomains regulate the distribution of signalling proteins, including β-adrenoceptors (βARs), and may act as membrane reserves, protecting cells during the mechanical stretch. Phosphodiesterase 4 (PDE4), particularly subtypes PDE4B and PDE4D, have been shown to localise in caveolae-rich membranes, where they tightly regulate βAR-induced cAMP production. Regional differences in membrane structure and βAR signalling also exist across the myocardium. This study investigates the role of caveolae in βAR-signalling in response to membrane deformation, the specific contributions of PDE4B and PDE4D, and regional differences in this regulation. We further assess how this signalling is altered in the context of heart failure. To induce membrane stretch-like stimuli, we used osmotic swelling in left ventricular cardiomyocytes isolated from healthy and failing (16 weeks post-myocardial infarction) rat hearts, as well as from human donor and failing (dilated cardiomyopathy) hearts. The βAR response was measured using a Förster Resonance Energy Transfer (FRET) reporter for the second messenger cyclic adenosine monophosphate (cAMP) and using CytoCypher for the measurement of cell contractility. Our study reveals a stretch-regulation of the β2AR signalling pathway that is dependent on functional caveolae and PDE4 activity, particularly in the basal region of the myocardium; this regulation is lost in heart failure. These findings indicate that caveolae are mechanosensitive membrane domains that undergo structural and functional changes in response to membrane deformation, thereby mediating the mechanical regulation of caveolae-associated signalling pathways.Open Acces
Imaging the structural connectome with hybrid MRI-microscopy tractography
No full textMapping how neurons are structurally wired into whole-brain networks can be challenging, particularly in larger brains where 3D microscopy is not available. Multi-modal datasets combining MRI and microscopy provide a solution, where high resolution but 2D microscopy can be complemented by whole-brain but lowresolution MRI. However, there lacks unified approaches to integrate and jointly analyse these multi-modal data in an insightful way. To address this gap, we introduce a data-fusion method for hybrid MRI-microscopy fibre orientation and connectome reconstruction. Specifically, we complement precise “in-plane” orientations from microscopy with “through-plane” information from MRI to construct 3D hybrid fibre orientations at resolutions far exceeding that of MRI whilst preserving microscopy's myelin specificity, resulting in superior fibre tracking. Our method is openly available, can be deployed on standard 2D microscopy, including different microscopy contrasts, and is species agnostic, facilitating neuroanatomical investigation in both animal models and human brains
Model-based geostatistical mapping of the prevalence of Onchocerca volvulus in Cameroon between 1971 and 2020
No full textBackground
After the closure of the African Programme for Onchocerciasis Control (APOC) in 2015, the Ministry of Public Health of Cameroon has continued implementing annual community-directed treatment with ivermectin (CDTI) in endemic areas. The World Health Organization has proposed that 12 countries be verified for elimination (interruption) of transmission by 2030. Using Rapid Epidemiological Mapping of Onchocerciasis, a baseline geostatistical map of nodule (onchocercoma) prevalence had been generated for APOC countries, indicating high initial endemicity in most regions of Cameroon. After more than two decades of CDTI, infection prevalence remains high in some areas. This study aimed at mapping the spatio-temporal evolution of Onchocerca volvulus prevalence from 1971 to 2020 to: i) identify such areas; ii) indicate where alternative and complementary interventions are most needed to accelerate elimination, and iii) improve the projections of transmission models.
Methodology
A total of 1,404 georeferenced (village-level) prevalence surveys were obtained from published articles; the Expanded Special Project for Elimination of Neglected Tropical Diseases portal for Cameroon; independent researchers and grey literature. These data were used together with bioclimatic layers to generate model-based geostatistical (MBG) maps of microfilarial prevalence for 1971–2000; 2001–2010 and 2011–2020.
Principal findings
Time-period was negatively and statistically significantly associated with prevalence. In 1971–2000 and 2001–2010, prevalence levels were high in most regions and ≥60% in some areas. Mean predicted prevalence declined in 2011–2020, reaching <20% in most areas, but data for this period were sparse, leading to substantial uncertainty. Hotspots were identified in South West, Littoral and Centre regions.
Conclusions/Significance
Our results are broadly consistent with recent MBG studies and can be used to intensify onchocerciasis control and elimination efforts in areas with persisting transmission, providing spatio-temporal prevalence trends to which transmission models can be fitted to improve projections of onchocerciasis elimination by 2030 and beyond
Defining the value proposition in diagnostic technology: challenges and opportunities for its understanding and development – a review with a multiperspective reflective analysis
No full textBackground: The Value Proposition (VP) in diagnostic technology serves as a “positioning statement” outlining the unique benefits, costs, and differentiation an innovation under development offers to healthcare organizations and its ability to effectively deliver these advantages in comparison to current interventions in the market. Despite its significance however, VP lacks a universally accepted definition, which is compounded by the diversity of technologies, their applications, and the varying needs of stakeholders. This paper aims to address this gap by offering a detailed conceptual analysis, revised definition of VP, and actionable recommendations for advancing VP development.
Methodology: We conducted a targeted narrative review, focusing on literature explicitly defining VPs in diagnostic technologies. Using Ovid’s Medline and Embase databases, we identified 19 relevant papers, of which only 5 provided explicit VP definitions. Our analysis incorporated principles of team science, encompassing reflective and thematic analyses of (1) interdisciplinary co-author discussions enabling us to weave together diverse insights into a cohesive exploration of the topic, and (2) MTech’s publicly available set of anonymised responses from NHS Associates, to capture the perspectives of the decision-makers and further enhance depth and breadth of our discourse.
Results and discussion: Our findings highlight the multifaceted nature of VP and its primary hurdles: inadequate identification of unmet needs and insufficient recognition of key stakeholders. We synthesized the evolution of VP definitions and explored the importance of unmet needs in their development, guided by frameworks, such as the Health Technology Navigation Pathway Tool, to ensure VPs meet both the pragmatic and aspirational goals of the healthcare. Thematic insights revealed opportunities for addressing these barriers through implementation science and collaborative strategies. This multi-perspective approach provided a conceptual examination of VP, enabling integration of varied viewpoints and insights.
Conclusion: By employing team science principles and reflective analysis, we introduced a revised definition of VP and a set of actionable recommendations to guide VP development in diagnostics. These findings highlight the importance of addressing stakeholder diversity, unmet needs, and the intricacies of blending interdisciplinary perspectives to advance the field
Validity of LEFM to measure the Mode II and Mixed Mode I/II fracture toughness of adhesively bonded CFRP
No full textIn this work, the validity of Linear Elastic Fracture Mechanics (LEFM) for measuring the energy release rates for adhesively bonded carbon fibre reinforced plastic (CFRP) joints under Mode II and Mixed Mode I/II loading, using end loaded split (ELS) and fixed ratio mixed mode (FRMM) specimens respectively, has been investigated. The G values determined using LEFM coupled with an effective crack length approach have been compared with the J-integral values measured simultaneously using the crack length independent slope-based J-integral methods. It is shown that higher GIIc values than JIIc values were measured using ELS specimens, but the GI/IIc values were in good agreement with the JI/IIc values measured using FRMM specimens. It is shown (by experimental and numerical investigations) that LEFM became invalid in Mode II due to the occurrence of local damage in the bondline of the specimen close to the clamp. This local damage resulted in erroneously high GIIc values being measured via LEFM. In contrast, the J method was able to provide accurate and valid toughness values as the contribution of local damage could be excluded through careful selection of the integral contour. Further analysis indicates that, if the local damage was eliminated by adding additional constraint to the specimen at the clamp, then valid values of GIIc via LEFM could be obtained. The J-integral method in this work provides an alternative tool to determine the fracture toughness of adhesive joints under Mode II loading in case of additional (secondary) damage outside the fracture process zone (FPZ) occurs, which is quite challenging for LEFM methods to separate the energy dissipated in and outside the FPZ