36750 research outputs found
Sort by
Research as an Intense and Inconvenient Gift: Participatory Youth Research and the Co-Productive Imagination
Get PDFThe article presents the research as a gift and an invitation to expand our imaginations as to how we are moved and seek to move others through research co-production. Thinking with the research as gift is made possible through a dialogue between Whitehead’s process philosophy (actual entities, societies, intensity, and attention) and Lauren Berlant’s queer and affect theory (inconvenience, friction, and attention). Research as gift is an intense, inconvenient, and attentional research practice. It is a practice that can be complex and fraught, yet this is what makes it so generative. The article demonstrates how a carousel of creative methods and organizational arrangements with sufficient expertise and capacity enable young people to co-produce research through eventful and expansive adventures following gift intensities and achieving life-affirming encounters for those inquiring in a world in process
Advancement In Skin Lesion Analysis Using Deep Learning Methods
Get PDFSkin cancer represents a major global health challenge, with steadily increasing incidence rates and substantial economic and clinical burdens associated with diagnosis and treatment. Among the various forms of skin
cancer, melanoma is the most aggressive and life-threatening, accounting
for a disproportionate number of skin cancer related deaths. Early detection
is critical, as melanoma is highly treatable in its early stages but rapidly becomes fatal once it progresses. These challenges motivate the development
of efficient, accurate, and scalable diagnostic tools capable of supporting
clinical decision making. In this context, deep learning has emerged as a
promising approach for automated analysis of dermatological images.
This thesis investigates the application of deep learning techniques for
the classification, generation, and segmentation of skin lesion images, with
a particular focus on melanoma. The overarching objective is to improve
model performance and reliability by examining key factors in dataset design and preprocessing, including data curation, synthetic data generation,
and the influence of lesion background on classification. By addressing
these factors, this work aims to enhance both the robustness and generalizability of deep learning models for skin cancer analysis.
The first part of the thesis examines the effect of duplicate and nearduplicate images on skin lesion classification using convolutional neural
networks (CNNs). Publicly available datasets, such as those from the ISIC
archive, are widely used in dermatological image analysis but often contain
redundant images that can bias training and inflate reported performance.
Multiple CNN models are trained using datasets curated with different duplicate removal threshold indices to systematically evaluate the impact of
redundancy. The results demonstrate that removing duplicate images leads
to more reliable performance estimates and improved generalization, highlighting the importance of careful dataset curation in medical image analysis.
The second part of the thesis focuses on synthetic dermatological image generation using generative adversarial networks (GANs).
Data scarcity, particularly for rare skin conditions and underrepresented
melanoma subtypes, remains a major limitation in training robust deep
learning models. To address this challenge, this research proposes a Combined Similarity Metric (cSM) for the evaluation and optimization of GANgenerated skin lesion images. The proposed metric provides a more informative assessment of synthetic image quality than conventional evaluation
approaches. In addition, a pseudo-label–based evaluation strategy is introduced to assess the realism and utility of synthetic melanoma images,
demonstrating their potential to enhance classification performance when
incorporated into training datasets.
The third study investigates whether clinically relevant attributes,such
as patient age, gender, anatomical lesion location, and lesion class are implicitly encoded within dermoscopic images. A multi-task learning framework is employed to predict these attributes simultaneously from a single
image. The findings suggest that while certain attributes can be inferred
with moderate accuracy, others remain challenging, revealing both limitations in current deep learning representations and opportunities for future
model and dataset improvements.
The final part of the thesis addresses skin lesion segmentation and
its influence on melanoma segmentation and classification. A novel
segmentation-based framework, DESeg SkinNet,is developed for segmentation efficiency and accuracy, and to evaluate the impact of background removal on classification performance. This study highlights the importance
of efficient lesion segmentation and optimized preprocessing pipelines in
improving downstream classification results.
Overall, this thesis contributes to the field of medical image analysis
by providing a comprehensive investigation into the role of data curation,
synthetic data generation, and segmentation in deep learning–based skin
cancer diagnosis. The findings underscore the importance of high quality
datasets and tailored preprocessing strategies for achieving reliable performance. The proposed methods demonstrate strong potential for improving
early melanoma detection and offer valuable insights for the development
of more robust and clinically applicable diagnostic systems
Co-Designing Accessible Trial Information: Lessons from Designing an Inclusive Patient Information Leaflet in the RaCeR2 Study
No full textBackground:
Informed consent is fundamental to ethical research, yet participant information leaflets (PILs) are often technical and difficult to understand. Although the importance of accessible study materials is widely recognised, practical guidance on how to achieve remains limited. For RaCeR 2, a randomised controlled trial evaluating different approaches to rehabilitation after shoulder rotator cuff repair, the baseline PIL was developed using Health Research Authority guidance. Early PPI consultation identified major concerns including dense formatting, overly complex language, and an unwelcoming tone, indicating the need for substantial redesign to support informed participation.
Description of Patient and Public Involvement Activity:
A diverse PPI group (n=5) supported the development of an accessible PIL. Contributors varied in age, gender identity, ethnicity, preferred language, employment status, disability, and experience of shoulder surgery. Engagement methods were tailored to participant needs and included online discussions, written feedback, and in-person “think-aloud” sessions. Given the depth of insight provided during initial PPI consultation, we adopted an iterative, user-centred approach drawing on co-design principles to enable contributors to directly influence the content, structure, and presentation of the PIL.
Outcome of the Patient and Public Involvement Activity:
Contributors identified challenges with the original materials, including confusing layout, inaccessible formatting, technical terminology, and a tone that did not feel supportive. Through iterative cycles of review and refinement guided by co-design principles, the leaflet was substantially redesigned to include clearer headings, formatting aligned with British Dyslexia Association guidance, bullet-pointed and tiered information, simplified explanations of data protection, and a more conversational tone. Accessibility testing confirmed compatibility with freely available online screen-reading software. During the subsequent regulatory review, inconsistencies between legal requirements and participant preferences highlighted tensions that may limit the accessibility of study materials.
Conclusions:
This commentary provides a pragmatic example of integrating co-design principles within PPI to create a more accessible PIL. Early engagement, flexible methods, iterative feedback, and testing with screen reading software were central to the process. Our experience also highlights the need for continued dialogue between researchers and regulators to ensure that participant-facing materials meet ethical and legal requirements while remaining understandable to all potential participants
Expressivity and corpus linguistics
No full textThis chapter explores how corpus linguistics can be used to analyze elements of expressivity in spoken discourse. Methods for transcribing and analyzing corpora for the study of expressive language are evaluated in terms of their strengths and limitations. The chapter discusses methods related to consistency and accuracy in transcription, narrow and broad transcription conventions, corpus annotation processes (including the use of AI), form-to-function and function-to-form analyses, generalizability of results, and selection of appropriate expressive features for corpus research. A case study examining the frequency and function of apologies uttered by American English speakers across various ages, genders, and regions is presented. This case study highlights the above methodological choices and illustrates how each choice may impact the findings of a corpus linguistic study of expressive language
The Temporal Limits of Single Positivity: Single Women in Postfeminist Culture
No full textThis article evaluates the extent to which the recent ‘single positivity movement’ has reconfigured representations of single womanhood in the UK. Drawing on in-depth interviews with 19 heterosexual women, aged 21–52, I argue that, in a postfeminist and neoliberal culture where heterosexual women have been encouraged to ‘dump him’ and prioritise their self and happiness, to free themselves from heteropatriarchal constraints, single womanhood has been transformed into an identity characterised by self-discovery, self-prioritisation, and self-love. In light of this movement, I find that some heterosexual women have become defiantly single and are unwilling to settle for the sake of settling down, choosing only to revoke their single status should a partner contribute to their already fulfilled lives. However, for most women in this study, while singlehood offered them a period of liberation to find themselves, prioritise themselves, and love themselves, this was dependent on their life-course stage, with an apparent temporal limit to single positivity. Therefore, despite a pervasive narrative of single positivity, which has seemingly facilitated a newfound acceptance to single womanhood, this article concluded that while singlehood is celebrated, and encouraged, as a temporary phase for women in their twenties, the couple norm is maintained as the dominant social form
14. The Invisibilisation of Male Victims in the Family and Criminal Courts: Domestic Abuse, Honour-Based Abuse, and Parental Alienation.
No full textHighlighting experiences and voices historically overlooked, misunderstood, or forgotten, this book answers the urgent need for a more diverse and inclusive curriculum in family law in England and Wales
Cybernetic Portraits at the Collision of Program and Problem: How can we draw diagrams of complex systems?
No full textSocial inclusion of people with severe mental illness: a review of current practices, evidence and unmet needs, and future directions
Get PDFSocial inclusion means being able to participate in activities valued within one’s community or wider society as one would wish. People with severe mental illness (i.e., psychoses, bipolar disorder, and severe depression) experience some of the highest rates of social exclusion compared to people with other disabilities. This is the case regardless of the availability of specialist mental health services. Therefore, questions arise about the extent to which mental health services can and do prioritize social inclusion as a goal of service provision, and what strategies are needed outside of mental health services, at the levels of legislation and policy, statutory services, and civil society. In this paper we consider what social inclusion means in different cultures and contexts, since the value attached to different activities varies by culture and by life stage and gender. We discuss the subjective impact of low levels of social inclusion in terms of loneliness, and the evidence base for interventions to address it. We then turn to strategies to increase observable forms of social inclusion, considering them at the levels of legislation, services and other community assets. While evidence for some interventions is largely based on the Global North, we use evidence and examples from the Global South to the extent that we have found them. We also consider the predominant frameworks for social inclusion used in health services, followed by alternatives that may offer a more empowering approach to social inclusion for some people. We then describe strategies to reduce social exclusion through interventions to address stigma and discrimination, directed at key target groups or at population level. We make recommendations for policy makers, researchers, health professionals, and advocates based on the evidence and examples we have found, covering various forms of legislation, services and mental health research. Our conclusions identify the next steps for interventions, including development, evaluation, implementation or modification for better contextual adaptation
Enhanced etching resistance of Y2O3 films through microstructure control via thermal annealing
Get PDFY2O3 films with enhanced etching resistance were fabricated on silicon substrates using RF magnetron sputtering followed by vacuum annealing at 300 °C, 600 °C, and 900 °C. The films were systematically evaluated under chemical, mixed, and physical plasma etching environments using CF4/Ar/O2 gas mixtures. Annealing at 900 °C transformed the microstructure from quasi-circular grains to densely packed polygonal grains with layered stacking, significantly improving the structural compactness of film. Y2O3 films demonstrated superior etching resistance compared to conventional materials including Al2O3, SiC, AlN, and quartz. The 900 °C-annealed film under vacuum showed highest resistance under chemical and mixed etching, while the 600 °C-annealed film under vacuum performed best under physical etching due to superior hardness (7.2 ± 2.0 GPa). Nanomechanical testing revealed that physical etching resistance correlates with hardness, while chemical and mixed etching resistance depends on the structural compactness of film. Nano-indentation and nano-scratch testing showed that coatings annealed in air were softer and less scratch-resistant than those annealed in vacuum, with differences being more pronounced in statistically distributed nano-scratch testing. X-ray photoelectron spectroscopy confirmed distinct mechanisms: chemical etching involves protective Ysingle bondF bond formation, mixed etching shows enhanced Ysingle bondF stability, and physical etching induces oxygen vacancies. This study establishes a cost-effective strategy for enhancing etching resistance through microstructural optimization, offering promising implications for semiconductor manufacturing applications
New perspectives on head and neck allometry and ecomorphology in tetrapods
Get PDFThe skull and neck are vital parts of the body, influencing feeding ecology, habitat exploitation and locomotion. Numerous studies have therefore sought to understand how the size of these segments vary with ecology and scale with overall body size. However, across past literature many different metrics have been used to represent both head and neck size, alongside disparate methods for body size normalisation and varied statistical approaches to analysing patterns. Furthermore, while several studies have examined allometric patterns across species of birds and dinosaurs, there are relatively few studies of other groups like mammals, non‐avian reptiles and amphibians. It is therefore currently difficult to combine or compare analyses from past studies to arrive at a clear picture of ecological and taxonomic trends in tetrapod head and neck allometry and evolution. To address these issues, we present a new analysis of head and neck proportions using a data set of 410 three‐dimensional digital skeletons that samples a wide taxonomic breadth of extinct and extant terrestrial tetrapods. Allometric and ecological patterns in head and neck size were analysed using phylogenetically informed approaches, with head and neck size quantified using multiple metrics representative of a range of methods used across previous studies. We find that different measurements used in the literature to represent head and neck size do not always yield qualitatively consistent results in terms of allometric patterns within and between major taxonomic and ecological groups. For example, across tetrapods, all metrics suggest negative allometry in skull size, whereas the pattern of allometry seen in the neck is influenced by the metric used (length versus volume). We also find that allometric patterns in linear metrics for head and neck size are better described by a linear model, whilst volumetric measurements are better fitted by a quadratic model for both the head and neck. Statistical support for quadratic models appears to be driven by species over 100 kg tending to show greater negative allometry in skull volume, whilst the neck shows strong positive allometry above this approximate size threshold. The disparate allometric patterns given by different metrics typically result from systematic variation in segment shape, which may often have adaptive significance. For example, distinct allometric trends in skull length and width are recovered across taxonomic and trophic groups, which may represent mechanical interactions between bite force and velocity in different feeding modes, particularly carnivory, insectivory and piscivory. Potentially adaptively significant patterns were also recovered in neck allometry for piscivores, and in the neck–head scaling seen in carnivores and herbivores, where a larger head in bigger carnivores enables capture of large prey but necessitates a reduced neck, while relatively smaller head sizes in herbivores (as food processing shifts to the gut) allow longer necks to increase the three‐dimensional volume (‘feeding envelope’) accessible to the head–neck system. The disparate qualitative and quantitative allometric relationships given by different metrics suggests that future work should carefully consider the choice of parameter used to represent skull and neck size when comparing trophic and taxonomic groups and making ecological and macroevolutionary inferences