26207 research outputs found
Sort by
AI under attack: Metric-driven analysis of cybersecurity threats in deep learning models for healthcare applications
Get PDFIncorporating Artificial Intelligence (AI) in healthcare has transformed disease diagnosis and treatment by offering unprecedented benefits. However, it has also revealed critical cybersecurity vulnerabilities in Deep Learning (DL) models, which raise significant risks to patient safety and their trust in AI-driven applications. Existing studies primarily focus on theoretical vulnerabilities or specific attack types, leaving a gap in understanding the practical implications of multiple attack scenarios on healthcare AI. In this paper, we provide a comprehensive analysis of key attack vectors, including adversarial attacks, such as the gradient-based Fast Gradient Sign Method (FGSM), evasion attacks (perturbation-based), and data poisoning, which threaten the reliability of DL models, with a specific focus on breast cancer detection. We propose the Healthcare AI Vulnerability Assessment Algorithm (HAVA) that systematically simulates these attacks, calculates the Post-Attack Vulnerability Index (PAVI), and quantitatively evaluates their impacts. Our findings revealed that the adversarial FGSM and evasion attacks significantly reduced model accuracy from 97.36% to 61.40% (PAVI: 0.385965) and 62.28% (PAVI: 0.377193), respectively, demonstrating their severe impact on performance, but data poisoning had a milder effect, retaining 89.47% accuracy (PAVI: 0.105263). The confusion matrices also revealed a higher rate of false positives in the adversarial FGSM and evasion attacks than more balanced misclassification patterns observed in data poisoning. By proposing a unified framework for quantifying and analyzing these post-attack vulnerabilities, this research contributes to formulating resilient AI models for critical domains where accuracy and reliability are important
Understanding adolescent girls’ thoughts and opinions on having social media influencers deliver body image and mental health support: A mixed-methods study
Get PDFObjectiveInfluencers have been used to deliver messaging in public health campaigns and have begun collaborating with researchers to deliver evidence-based content. Building on this approach, this study explored adolescent girls’ perspectives on influencers and utilising influencers to deliver body image and mental health interventions.MethodsA mixed-methods study was conducted with 375 UK-based girls (Mage = 15.39, SD = 1.63) to understand their experiences with influencers and their opinions about influencer-delivered body image and mental health interventions. Data were analysed quantitatively via descriptive statistics and qualitatively using inductive content analysis.ResultsMost participants expressed likeability, credibility, authenticity, and relatability were key influencer qualities when choosing to trust an influencer, whilst their ethnicity, gender, and body size were not important. In considering intervention development, the most salient factor in trusting an influencer about body image was that the influencer shared their own experiences. Participants preferred offline body image support (47%) versus online (38%); however, just over 60% agreed that influencer-delivered body image interventions were a good idea. The content analysis generated three main categories: ‘Influencer-delivered interventions would be helpful’ (53.5%), encompassing the offer of support and online environment as ideal; ‘influencer-delivered interventions would be unhelpful’ (28.1%), including influencers are not qualified and the toxicity of the internet; and ‘influencer qualities’ (18.3%), such as relatability.ConclusionOverall, adolescent girls thought an influencer-driven approach could be helpful in improving their body image and mental health. Caveats included that the influencer must be trustworthy, relatable, and authentic, which could encompass self-disclosure related to their mental health
From classroom to clinic: Innovating radiotherapy treatment planning education through real-world end-to-end case study simulation with an anthropomorphic phantom
Get PDFThe incorporation of simulation-based learning in healthcare education, particularly in radiotherapy, is necessary for enhancing training and professional competencies to serve patient safety and treatment accuracy. This study aimed to incorporate an innovative end-to-end case study methodology, utilizing an anthropomorphic head phantom, into an undergraduate radiotherapy program at a United Kingdom (UK) based university. The objective was to enhance students' practical learning and theoretical understanding in radiotherapy treatment planning, a field where precision and accuracy are paramount. The study began with an exploratory literature review to identify key educational challenges and opportunities in radiotherapy treatment planning. A qualitative approach was employed, using a focus group methodology to gather in-depth insights from subject experts, including educational and clinical professionals involved in undergraduate radiotherapy teaching. The focus group discussions explored the integration of an anthropomorphic head phantom within a simulated, case study-based training framework. This innovative approach combined practical skills development with theoretical learning, promoting active engagement and mirroring real-world clinical scenarios. Focus group discussions showed favorability towards the end-to-end case study method in simulation-based learning. Participants emphasized evaluating plans through assessments and using supplementary tools like video guides and workbooks to enhance learning. Incorporating the anthropomorphic phantom marked a notable advancement, offering authentic training possibilities in radiotherapy undergraduate education. The study demonstrates the potential of integrating an end-to-end teaching concept in radiotherapy education. By providing a realistic and comprehensive training experience, the approach can further enhance student engagement and learning outcomes. While real-world testing is pending, this innovative methodology shows promise in shaping proficient future radiotherapy graduates, highlighting the need for continuous evolution in educational strategies to meet the demands of modern healthcare training. [Abstract copyright: © 2025. The Author(s).
Written evidence to the public accounts committee’s inquiry into local government financial sustainability
No full textMy evidence relates to the subject of: Support for financially failing local authorities and action to tackle systemic issues. Within this, I focus specifically on the role of POCA and Accredited Financial Investigators (AFIs) in generating income for local authorities. It is notable that neither are mentioned anywhere in the NAO’s 2025 overview report on local government funding and spending. This oversight is indicative of a broader trend which sees POCA and AFIs frequently overlooked or misunderstood, despite their significant potential to generate income, as I will outline below
Complete genome of five Acinetobacter baumannii phages
Get PDFAcinetobacter baumannii tailed phages Bhz9, Bhz15, Bhz16, Bhz20, and Bhz21 were isolated from wastewater samples in Hamburg-Eppendorf (Germany). Their double-stranded DNA genomes range from 41,054 to 167,277 bp, encoding between 54 and 267 predicted genes, with a G + C content ranging from 36.4% to 48.9%. Only phage Bhz15 encodes tRNAs
Explore the relationship between house prices and crime rate in the UK using machine learning techniques
Get PDFIn the UK, house price estimation is an important topic. It has been discussed in numerous academic research publications and official and business reports. A house’s price can change depending on size, age, and location. One of the main characteristics of a property’s neighbourhood is the crime rate; it is crucial to consider how it may affect the home’s price. If crime rates rise, a neighbourhood may lose appeal to potential purchasers in favour of comparable-priced neighbourhoods. This study employs data from January 2020 to April 2022 to examine the relationship between crime rates and house prices in Bristol, UK. Crime datasets were taken from the UK Crime Stats website, while data on house prices was taken from the Price Paid Data of the HM Land Registry. 34,000 transaction records and criminal statistics are combined in the study to provide a comprehensive dataset for analysis. Two tree-based machine learning algorithms-Decision Tree (DT) and Random Forest (RF)-and Exploratory Data Analysis (EDA) were used to model and evaluate the data. The results show that the Random Forest outperformed the Decision Tree regarding prediction accuracy. According to the results, there is a substantial correlation between crime rates and house prices. While vehicle crimes and bike theft had a positive correlation with property values, violent crimes had the opposite effect. These revelations highlight the intricate relationship between different kinds of crime and the characteristics of the housing market, with considerable consequences for investors, policymakers, and real estate developers
The Crown Dependencies and Overseas Territories: A comparative constitutional and political analysis
Get PDFIn the realm of the United Kingdom (UK) there are three Crown Dependencies (CDs) and fourteen Overseas Territories (OTs). The UK has ultimate responsibility for them, but the territories have significant autonomy. However, it is unusual for the territories to be examined together. There are several reasons for this, including differences in how they became associated with the UK, their distinctive constitutional and political arrangements, and their varied geographical locations. The article not only highlights these differences but also suggests where the two sets of territories are more aligned than first thought. Issues such as colonialism, constitutional conventions, and legislative decision-making are considered, so too institutions including the Governor, Whitehall, and UK Parliament's committee system. The article contends there are interesting and unseen similarities, and there are possibilities for the OTs to close the gap in governance to better align with the CDs and their relationship with the UK
Trauma has an echo: A qualitative study of barriers to routine healthcare for survivors of sexual violence in UK higher education
Get PDFObjectives: To explore barriers to healthcare among survivors of sexual violence (SV) and the behavioural pathways behind avoidance, focusing on survivor-led solutions. Design: A mixed methods study collated qualitative interviews/surveys to explore the lived experiences of survivors of SV. Data were analysed using both quantitative descriptors and qualitative thematic analysis to expand the mechanisms/solutions to reported rates. Setting: Higher education setting in the UK. Participants: Forty-two survivors of SV between the ages of 18 and 29 self-identified as female (36), male (4), genderfluid (1) and non-binary transmasculine (1), with 10 describing themselves as being from racially minoritised communities and 32 as White survivors. Results: Analysis found 86% of survivors completely or significantly avoided healthcare, particularly sexual and reproductive services. Three themes were identified: (1) wider societal blame/marginalisation of survivors hindered their ability to access care in what felt like ‘a system of oppression’; (2) once past these barriers, direct experiences with professionals replicated trauma, exacerbating avoidance and health disparities through ‘healthcare-induced re-traumatisation’; (3) survivors identified what they needed to re-engage in healthcare including trauma-informed professionals and compassionate services with ‘survivor-centred care’. Conclusions: SV may deepen health inequalities as survivors avoid healthcare. Survivor-led reforms called for survivor-centred practices and encouraged systemic reflection on how healthcare systems may contribute to the broader marginalisation of survivors. Findings echo policy recommendations for co-produced services led by minoritised/marginalised patients and operationalise trauma-informed training for healthcare professionals. Additionally, access-focused psychological support could reduce the impact of sexual trauma on morbidity and mortality
Body image among people living with HIV: A systematic review
No full textThe aim of this systematic review was to analyse, synthesise and review existing research on body image among people living with HIV (PLWH). Specifically, the review investigated the relationship of body image with sociodemographic and clinical variables and psychological correlates among PLWH. A comprehensive literature search was conducted using Web of Science, PsycARTICLES, MedLine, ProQuest and Scopus databases and yielded a total of 699 abstracts. The inclusion criteria encompassed quantitative and peer-reviewed articles published in English devoted to HIV/AIDS adult patients, with validated measurements of body image operationalized explicitly as an outcome variable. A total of 17 articles were accepted for systematic review and analysis. The quality of the studies was evaluated using STROBE guidelines for observational studies. The findings revealed that body image disturbances are common among PLWH and are influenced by factors such as lipodystrophy, depression, anxiety, self-efficacy in health behaviours, social support, stigma and quality of life. Higher body image disturbance was also linked to poorer adherence to antiretroviral therapy (ART) and increased HIV sexual transmission risk behaviours. Body image is a crucial aspect of the psychosocial functioning of PLWH. Despite medical advancements, PLWH continue to face significant body image disturbances that impact their quality of life. Future research should focus on longitudinal studies, diverse geographical contexts and interventions aimed at improving body image among this population. Addressing body image concerns through psychosocial support and therapeutic interventions can potentially enhance the overall well-being and health outcomes of PLWH. [Abstract copyright: Copyright © 2025. Published by Elsevier Ltd.
Forskning i egen praksis – in indføring for praktikere
No full textAimed at practicing psychotherapists, psychotherapy students, and other relevant professional groups in psychosocial fields, this book serves as a guide into research within one's own practice and explores the importance and methods involved. It introduces the concept of the practitioner as a researcher and key areas covered include the purpose and importance of practitioner research, addressing barriers that hinder practitioners from seeing themselves as researchers, and exploring different types of research, notably the distinction between 1st-, 2nd-, and 3rd-person research, which is highlighted as a crucial tool for a necessary mental shift for practitioners. The book provides guidance on various data collection methods applicable to practitioner research. It also outlines different data analysis strategies, including inductive, deductive, and abductive approaches, emphasizing analytical transparency. Significant attention is given to ethical considerations, delving into challenges and responsibilities practitioners face, especially when involving clients, and stressing the importance of informed consent and handling potential dual roles. The book discusses Action Research and the use of "trial actions" as a cyclical process for implementing and evaluating changes. Approaches like Phenomenology, focusing on understanding lived experience through methods like bracketing and descriptive language, and Autoethnography, using personal experience to understand broader phenomena, are also explored as relevant methodologies, and the value of combining qualitative and quantitative methods in Mixed Methods Research is discussed. The book emphasizes the importance of genre awareness for reporting practitioner research, requiring methodological trustworthiness and analytical transparency