Open Research Exeter - University of Exeter
Not a member yet
41213 research outputs found
Sort by
Plasmonic Nanosensors Embedded in Nanocrystalline Cellulose Hydrogel for the Detection of Reactive Oxygen Species: Towards a Sensing Bandage
No full textEarly and accurate detection of biomarkers associated with oxidative stress is essential for monitoring chronic wound progression and guiding treatment decisions. Among reactive oxygen species (ROS), hydrogen peroxide (H₂O₂) plays a pivotal role as both a signalling molecule in healing and a marker of prolonged inflammation. Conventional detection methods for H₂O₂ often fall short in terms of sensitivity, selectivity, and real-time applicability in complex biological environments. Surface-enhanced Raman scattering (SERS), a powerful extension of Raman spectroscopy, offers ultrahigh sensitivity, molecular specificity, and the ability to function in aqueous, optically scattering media, making it an ideal candidate for in situ biochemical sensing of ROS. This thesis aimed to develop a tailorable SERS-based sensing platform for H₂O₂ detection, with a view towards future non-invasive wound monitoring. The first experimental chapter, Chapter 2, of this study focused on the preparation of SERS-active gold (Au) core-satellite nanoassemblies, assembled using hyperbranched polymers (HBP) polymerized through reversible addition-fragmentation chain transfer (RAFT) polymerization. These innovative polymers were designed to serve dual roles as a structural linker and Raman tag carrier. They were functionalized with epoxide groups for covalent attachment of aromatic thiols, which acted as Raman tags due to their high Raman cross-sections. The incorporation of distinct Raman tags in the Au nanoassemblies allowed generation of tuneable SERS barcodes, where each nanoassembly produced a unique and programmable spectral fingerprint suitable for future multiplexed sensing.Chapter 3 introduced a method for forming Au core-satellite nanoassemblies, in which a reactive probe was incorporated within the plasmonic hotspots between the core and satellite AuNPs. Luminol’s reactivity toward H₂O₂ was exploited, enabling its use as a molecular reporter for sensing H₂O₂. Unlike conventional applications relying on its luminescence, luminol was innovatively employed here as a reactive SERS probe, where its oxidative transformation generates distinct vibrational signatures detectable via SERS. The surface of the core AuNPs was first labelled with luminol, followed by the attachment of a RAFTsynthesized hyperbranched polymer (HBP) bearing alkyne and trithiocarbonate end groups, both of which have strong affinity for gold surfaces. This assembly approach produced a morphology suitable for direct SERS-based detection of H₂O₂, with characteristic spectral changes indicating reaction with H2O2. A key aim of this chapter was to evaluate the responsiveness and selectivity of the sensor. Kinetic studies were conducted across a range of pH values (7.4-10.0), identifying pH 9.3 as optimal for sensor activity and physiologically relevant to chronic wound environments. Importantly, the nanosensor demonstrated a linear SERS response to increasing H₂O₂ concentrations, highlighting its quantitative potential. Selectivity was validated by testing the nanosensor’s ability to specifically detect H2O2 in a biologically complex medium such as Fetal Bovine Serum (FBS), confirming its reliability in complex environments.Chapter 4 involved embedding the sensing Au nanoassemblies into a Nanocrystalline Cellulose (NCC) hydrogel to assess performance within a biocompatible matrix relevant to wound monitoring applications. The hydrogel preserved both its structural integrity and the SERS functionality of the embedded nanoassemblies. To further understand the sensing dynamics, analyte diffusion through the NCC hydrogel was experimentally characterized and modelled using the one-dimensional transient heat conduction equation, which serves as an analogue for the diffusion processes. This provided insight into reaction and transport behaviour relevant to in situ detection. The NCC hydrogel-embedded Au nanoassemblies also demonstrated feasibility for non-invasive, deep sensing using Surface Enhanced Spatially Offset Raman Spectroscopy (SESORS), with successful signal detection through the gel and a simulated wound dressing.This thesis presents a multidisciplinary platform that bridges polymer chemistry, plasmonic nanotechnology, and chemical engineering. The modular design enables chemical specificity, quantitative capability, and compatibility with realistic biological systems. These findings lay the groundwork for further ex vivo and in vivo validation, and future development of wearable or injectable sensors for continuous wound monitoring and personalized healthcare</p
Soluble HLA Class I Is Released From Human β-Cells Following Exposure to Interferon
No full textHLA class I (HLA-I) molecules present intracellular antigenic peptides to CD8+ T cells during immune surveillance. In donors with type 1 diabetes, hyperexpression of HLA-I occurs in islets with residual insulin-producing β-cells as a hallmark of the disease. HLA-I hyperexpression is frequently detected beyond the islet boundary, forming a “halo.” We hypothesized that this halo may reflect the diffusion of soluble forms of HLA-I (sHLA-I) from the islets to the surrounding pancreatic parenchyma. To verify this, we assessed the expression of total, cell surface, and sHLA-I in β-cell lines and isolated human islets after treatment with interferon-α (IFN-α) and IFN-γ. Consistent with the expression patterns of HLA-I in situ, the β-cell lines and cultured human islets dramatically upregulated total and surface HLA-I when exposed to IFNs. Concomitantly, sHLA-I release was significantly increased. HLA-I released within extracellular vesicles and cleaved forms of HLA-I did not significantly contribute to the sHLA-I pool. Rather, IFNs upregulated mRNA splice variants lacking the transmembrane domain. Our findings suggest that β-cells respond to IFNs by upregulating cell-associated and soluble forms of HLA-I. Soluble HLA-I may play a role in modulating islet inflammation during the autoimmune attack.</p
Drones in ecology: ten years back and forth
No full textA decade after our initial publication predicting that lightweight drones would revolutionize spatial ecology, drone technology has become firmly established in ecological studies. In the present article, we explore the key developments in ecological drone science since 2013, considering plant and animal ecology, imaging and nonimaging workflows, advances in data processing and operational ethics. Focusing on inexpensive, lightweight drones equipped with various sensors, we offer a critical evaluation of drone futures for ecologists, arguing that this could deliver opportunities for volumetric ecology to take flight. We discuss the potential future uses of drones in aerobiology and in understory and underground ecological studies and debate the future of multirobot cooperation from an ecological standpoint. We call on ecologists to engage critically with drone technology in this next phase of development.</p
We are storytelling apes: Experimenting with new scientific narratives in a time of climate and biodiversity collapse
No full textProvoked by a lack of appropriate political action on the global climate and biodiversity crisis, we present a perspective advocating and demonstrating a new plurality in scientific communication methods.Science writing, in being objective and dispassionate, actively seeks to mask empathetic connection and hides curiosities that may exist between authors and their subjects. We explain, referencing other work, why scientific writing in academic journal articles is problematic as a singular method of communication that fails to engage non-specialists. A wealth of existing philosophical work suggests that science translated into stories can deliver a range of valuable outcomes both within science and to wider society. However, such work may not be accessible to busy scientists (i.e. published in non-science journals, born from different epistemologies, using unfamiliar lexicon) and fails to make simple suggestions about how scientists can implement storytelling themselves.In this perspective, we outline the problems with existing means of engaging people with climate and environmental science, and review storytelling philosophy work, giving examples of how narrative has improved communication, transformed understanding and shifted opinion. We argue for scientific communication models that are beyond objective, both to engage people within science (i.e. across different epistemologies and disciplines) as well as outside of science (i.e. the public and policy makers). We suggest that work that enhances public engagement and trust in science is urgent and of paramount importance, particularly when the policies needed to effect change require huge shifts in behaviour.Crucially, we explain how the augmentation of scientific writing with more narrative forms need not compromise the objectivity of science. Differently from other storytelling in scientific work, we propose three possible ways to diversify environmental science communication and position this piece as a radical intervention into what has become a singularity of otherwise dispassionate scientific writing.</p
The birth of naval history: audience and objectivity in British eighteenth-century historical writing
No full textDuring the early decades of the eighteenth century the first general naval histories were published in Britain. Individuals from a range of backgrounds claimed ownership of this new and burgeoning subject, all professing to have produced the most reliable and accurate histories of the navy. Their competing publications provoked considerable debate, with authors drawn into fighting a war of words in the national press. These disputes raised important questions about the study of the past. Who should write naval history? How should it be written? Who was it written for? This article places naval history in the broader context of British politics, society and culture, investigating three works of naval history published between 1720 and 1735. It analyses their authorship and intended audience, suggesting that the production of these works was a symptom of a society quickly becoming enamoured with its naval past, and keen to record and disseminate this history. In addition, this article will explore the debates over objectivity, expertise and authority that were central to discussions over who should write naval history. In doing so, it contributes to a broader historiography that considers the development of the historical profession in the early eighteenth century.</p
Multi-class Network Intrusion Detection with Class Imbalance via LSTM & SMOTE
No full textMonitoring network traffic to maintain the quality
of service (QoS) and to detect network intrusions in a timely and efficient manner is essential. As network traffic is sequential, recurrent neural networks (RNNs) such as long short-term memory (LSTM) are suitable for building network intrusion detection systems. However, in the case of a few dataset examples of the rare attack types, even these networks perform poorly. This paper proposes to use oversampling techniques along with appropriate loss functions to handle class imbalance for the detection of various types of network intrusions. Our deep learning model employs LSTM with fully connected layers to
perform multi-class classification of rare network attacks. We enhance the representation of minority classes: i) through the application of the Synthetic Minority Over-sampling Technique (SMOTE), and ii) by employing categorical focal cross-entropy loss to apply a focal factor to down-weight examples of the majority classes and focus more on hard examples of the minority classes. Extensive experiments on KDD99 and CICIDS2017 datasets show promising results in detecting network intrusions (with many rare attack types, e.g., U2R, R2L, Probe, Infiltration,
Hearbleed, etc.).
Index Terms—Network Intrusion Detection, Deep Learning,
Rare Attacks, LSTM, Class Imbalance, SMOTE.</p
Is it a ‘Boys Club’ for a Reason? How American Presidential Politics are Masculine Biased in Character and Competency
No full textGender and politics research shows that there is an existing and sustained bias for masculine traits in American presidential politics. This manifests itself significantly in the perception of a candidate’s character and competency. Female presidential candidates are questioned in their qualification in these two areas predominantly, with greater focus and a pessimistic view comparatively to male candidates. The media is the biggest perpetrator in this pessimism, with coverage of presidential candidacies featuring an obvious and stereotyped bias towards male candidates. There are also drastically less viable female presidential candidates, as history has dictated their involvement as minimal or inadequate for political needs. This concentrates the media’s maltreatment.This work shows that candidates must be dominantly masculine in character and highly competent in masculine policy focuses, such as conflict and defence. An issue for women is that they are pigeonholed to stereotyped feminine policy. Masculinity and femininity connote different traits, and masculine traits are the preference. Therefore, female candidates are viewed as less presidentially characterful and less presidentially competent because they are seen to naturally lack the requirements. Bias is shown where machismo and aggression are praised, as well as in the novelisation and lack of campaign support for female candidates in comparison.Evidence shows that acceptance of feminine traits depends on the policy area, private lives, party identification, and events at the time of candidacy. It also depends on the gender stereotypes acknowledged by the voter, male or female. Victoria Woodhull in 1872 had significantly less voter and media support than Hillary Clinton in 2008 as expected, though research shows that areas of disqualification regarding character and competency were alarmingly similar. Thus, female candidates must carefully balance their femininity with expected masculinity in multiple contentious areas. In summary, the research highlights a continued and known bias towards male and masculine presidents.</p
Imposter
No full textThe Imposter comic brings the real-life experience of shame experienced by a medical student at a Signaporean medical school to life. Through a story that follows four different fictionalised characters, the comic creatively presents unique experiences of shame and reflects Singapore’s diverse culture. With its rich imagery and lan?guage, Imposter gives readers from around the world unique insights into the culture of a Singaporean medical school. As shame is an often unspoken and taboo experience for healthcare providers, this comic is an import?ant contribution to the medical humanities literature. The limited evidence available makes clear that shame is a common experience in medical training. Among both learners and professionals in healthcare, shame can lead to defensive medicine and moral injury; can impede learning, erode trust and empathy; can cause individ?uals to leave the profession; can cause under-reporting of errors, and can contribute to burnout and stress. By encouraging open dialogue and reflection on shame experiences, thus normalising and overcoming the stig?ma attached to these experiences, we can simultaneously improve healthcare workers’ experiences and patient safety. Imposter contributes to this important endeavour of normalising emotions in healthcare and helping shift the stigma and taboo around the emotion of shame.</p
A Systematic Review of Trainee and Qualified Clinical Psychologists’ Experiences of Self-Disclosure and Nondisclosure Inter-Professionally. An Exploration of How UK DClinPsy Selection Leaders Perceive Applicant Disclosure of Lived Experience During Selection.
No full textObjectives: To examine dominant discourses utilised by Doctorate in Clinical Psychology (DClinPsy) selection leads in the United Kingdom (UK) when discussing: perceptions of Lived Experience (LE) disclosure during selection, LE in the context of DClinPsy competency, and negotiation of conflicting perspectives of LE within selection teams.
Methods: 13 UK DClinPsy selection leads from 12 Universities participated in one-to-one semi-structured interviews. A Critical Discursive Psychology approach to analysis identified interpretative repertoires, ideological dilemmas, and subject positions.
Results: Analysis identified that LE is considered a ‘broad’ concept on a ‘spectrum’ or ‘continuum’. There was a repertoire of perceiving LE as a motivation for entering Clinical Psychology. However, LE requires reflection to be a strength in the profession. LE was perceived as an ‘asset’ that builds ‘reflection’, ‘contextual awareness’, and ‘compassion and empathy’ competencies. LE was constructed as one of many routes to developing these competencies. This presents a dilemma when participants discuss LE as bringing another ‘level’ to these skills. Selection leads identified shared values, training, and facilitating discussion as key repertoires within their leadership strategy. They consider the power and lack of power they have in their roles.
Conclusions: Reflection was a core repertoire for selection leads, highlighting the importance for Clinical Psychology selection to have a clear framework for understanding reflection quality. Repertoires suggest LE may be welcomed within DClinPsy selection, yet complex discussions within selection teams to manage these perspectives is still required. Limitations and areas for future research are discussed.</p
SEDyConv: Spatially enhanced multi-dimensional dynamic convolution for medical multi-organ segmentation in CTs
No full textAutomated multi-organ segmentation presents a considerable challenge owing to the diversity of organs and individual variations. Current state-of-the-art deep-learning techniques rely primarily on static kernel weights that are fixed after training, thereby limiting their flexibility in adapting to diverse inputs. In this study, we propose a novel multi-organ segmentation method using a plug-and-play three-dimensional dynamic convolution module. This method is designed to address the challenges posed by the variability of CT scans in contrast to static segmentation models. We uniquely leverage multiple input-dependent attention mechanisms to adjust the coefficients across four dimensions of convolutional kernels dynamically, offering enhanced adaptability. This approach surpasses traditional feature-based dynamic methods in terms of flexibility, which is attributable to the global sharing of kernel parameters and a smaller kernel shape. In addition, we utilize a refined local block to preserve the spatial properties and extend the convolutional kernel space to N dimensions, thereby efficiently enhancing the representational capabilities of the model through higher-dimensional feature fusion. Furthermore, we design dynamic switches to integrate multi-dimensional global and local information adaptively, guiding the model to generate feature maps that closely align with the input characteristics. Visualizations of the dynamic coefficients and features generated by different inputs clearly demonstrate the adaptability of our method. Extensive experiments on four multi-organ segmentation datasets with various labeled organs and scales indicate that our proposed method outperforms other state-of-the-art methods in terms of improving the segmentation accuracy, particularly for organs with complex morphologies or small sizes. Code available at: https://github.com/lihaoqin168/SEDyConv.</p