19200 research outputs found
Sort by
14th International Conference of the Financial Engineering and Banking Society
No full textThis paper reconciles the ‘mean-variance (MV) vs. equal-weighting (1/N)’ debate using an unbiased performance-to-error map and delineates thresholds where subjective interventions become essential to surpass MV optimisation. Traditional portfolio theory validations are often inconclusive due to a ‘joint-test’ problem entangling alpha and risk models. Using a Monte Carlo-based methodology that isolates portfolio selection, this paper offers an impartial assessment of the performance sensitivity of portfolio-selection models to estimation errors and clarifies when and why appropriate subjectivity can enhance performance beyond MV. The findings underscore the need to embed subjectivity and rational investor behaviour more systematically into portfolio theory and practice
ESA-YOLOv5m: a lightweight spatial and improved attention-driven detection for brain tumor MRI analysis
Get PDFIntroduction: The early and accurate detection of brain tumors is vital for improving patient outcomes, enabling timely clinical interventions, and reducing diagnostic uncertainty. Despite advances in deep learning, conventional Convolutional Neural Network (CNN)-based models often struggle with small or low-contrast tumors. They also remain computationally demanding for real-time clinical deployment. Methods: This study presents an Enhanced Spatial Attention (ESA)-integrated You Only Look Once v5 medium (YOLOv5m) architecture, a lightweight and efficient framework for brain tumor detection in MRI scans. The ESA module, positioned after the Spatial Pyramid Pooling-Fast (SPPF) layer, enhances feature discrimination by emphasizing diagnostically relevant regions while suppressing background noise, thereby improving localization accuracy without increasing computational complexity. Experiments were conducted on the Figshare brain tumor MRI dataset containing three tumor classes: glioma, meningioma, and pituitary. Results: ESA-YOLOv5m achieved a Precision of 90%, Recall of 90%, and mean Average Precision (mAP)@0.5 of 91%, surpassing the baseline YOLOv5m by approximately 11%–12%. An ablation study further confirmed that placing the ESA module after the SPPF layer yields the highest performance ([email protected] = 0.91), while earlier integration produced marginally lower results. Classwise analyses demonstrated consistent gains (mAP range 0.87–0.98), and fivefold cross-validation showed stable performance ([email protected] = 0.910 ± 0.006). Efficiency tests revealed negligible overhead, with less than a 4.3% increase in parameters and an average latency below 10 ms per image. Discussion: Overall, the results validate that integrating a lightweight spatial attention mechanism significantly enhances tumor localization and model generalization while preserving real-time inference. The proposed ESA-YOLOv5m framework provides a reliable and scalable solution for automated brain tumor detection, suitable for clinical decision-support systems and edge healthcare applications
Perinatal Imaging in Partnership with Families (PIPKIN): Longitudinal cohort study protocol
Get PDFWhile advances in behavioural and neuroimaging methods suitable for use with infants have greatly increased our understanding of infant brain function, cognition and behaviour in recent years, relatively little is known about the rapid period of development during the last trimester of pregnancy and first weeks and months after birth, as well as the roles that the social environment and stress play in shaping this development. This protocol paper outlines The UK Perinatal Imaging in Partnership with Families (PIPKIN) Study, a unique, multi-method, longitudinal cohort study investigating the early development of fetal and infant neurocognitive function and behaviour, and how the infant's social and family environment shapes this development. The study follows families from a range of socio-economic backgrounds who participate at ten timepoints, from the third trimester of pregnancy until their infant is nine months old, with three visits taking place during the infant's first postnatal month. The study harnesses recent methodological advances coupled with the drive for more ecologically valid data collection by undertaking many of these visits in families' homes. Methods include measures of fetal behaviour using 4D ultrasound scanning; infant brain imaging using fNIRS and EEG; a full-day video recording of the home environment from the infant's perspective, with physiological measures; measures of recent stress in both infant and mother; questionnaires relating to the home environment as well as parents' feelings, attitudes, health and parenting routines; and standardised measures of infant behaviour and development. Specific aims are to investigate: i) individual differences in basic sensory, behavioural and motor processing between late prenatal and early postnatal periods; ii) rapid change in cortical functions over the first month, particularly for brain networks that support social behaviour; iii) effects of social interaction on developing brain function; and iv) individual differences in developmental trajectories associated with poverty-related contextual factors
Mitigating Insider Threats: Insights from Software Security Experts for Process Improvement and Risk Reduction
No full textThe efficacy of coaches’ delivery of teaching personal and social responsibility in a youth sports club
No full textA consensus statement on child and family health during the COVID-19 pandemic and recommendations for post-pandemic recovery and re-build
Get PDFIntroduction: As health systems struggled to respond to the catastrophic effects of SARS-CoV-2, infection prevention and control measures significantly impacted on the delivery of non-COVID children's and family health services. The prioritisation of public health measures significantly impacted supportive relationships, revealed their importance for both mental and physical health and well-being. Drawing on findings from an expansive national collaboration, and with the well-being of children and young people in mind, we make recommendations here for post-pandemic recovery and re-build. Methods: This consensus statement is derived from a cross-disciplinary collaboration of experts. Working together discursively, we have synthesised evidence from collaborative research in child and family health during the COVID-19 pandemic. We have identified and agreed priorities areas for both action and learning, which we present as recommendations for research, healthcare practice, and policy. Results: The synthesis led to immediate recommendations grouped around what to retain and what to remove from “pandemic” provision and what to reinstate from pre-pandemic, healthcare provision in these services. Longer-term recommendations for action were also made. Those relevant to children's well-being concern equity and relational healthcare. Discussion: The documented evidence-base of the effects of the pandemic on children's and family services is growing, providing foundations for the post-pandemic recovery and re-setting of child and family health services and care provision. Recommendations contribute to services better aligning with the values of equity and relational healthcare, whilst providing wider consideration of care and support for children and families in usual vs. extra-ordinary health system shock circumstances
Corticospinal excitability is facilitated during coordinative action observation and motor imagery
No full textCombined action observation and motor imagery (AOMI) facilitates corticospinal excitability (CSE). This study used single-pulse transcranial magnetic stimulation (TMS) to explore changes in CSE for coordinative AOMI, where the observed and imagined actions are related but not identical, for a single-leg sit-to-stand (SL-STS) movement. Twenty-one healthy adults completed two testing sessions including baseline (BL), action observation (AO), and motor imagery (MI) control conditions, and three experimental conditions where they observed a slow-paced SL-STS while simultaneously imagining a slow- (AOMIHICO), medium- (AOMIMOCO), or fast-paced (AOMILOCO) SL-STS. A TMS pulse was delivered to the right leg representation of the left primary motor cortex at three stimulation timepoints aligned with peak electromyography (EMG) activity of the knee extensor muscle group for the fast- (T1), medium- (T2), and slow-paced (T3) SL-STS during each condition. Motor evoked potential (MEP) amplitudes were recorded from the knee extensor muscle group as a marker of CSE for all stimulation timepoints and conditions. A main effect for experimental condition was reported for all stimulation timepoints. MEP amplitudes were significantly greater for AOMIHICO at T1 and T3, and AOMIMOCO and AOMILOCO at all stimulation timepoints, when compared with control conditions. This study provides neurophysiological evidence supporting the use of coordinative AOMI.</p
Michael Shaw in conversation with Alexandra Kolb
Get PDFInterview conducted as part of the BA/Leverhulme-funded project 'On Air: Exploring the historical, scientific and political impacts of inflatables in choreography
