110830 research outputs found
Sort by
Staff experiences of percutaneous tracheostomy in intensive care : challenges, complications and potential solutions
Get PDFSummary: Tracheostomies are performed in 10–13% of UK intensive care admissions. While the percutaneous technique is well established, accurate needle insertion can be difficult in patients with obesity or neck swelling. These challenges increase the risk of bleeding, airway loss and injury to neck structures. In anatomically complex cases, patients are referred for surgical tracheostomy, causing delays and additional healthcare costs. Semi‐structured interviews and focus groups were conducted with 32 staff across three intensive care units. Participants included consultants, resident doctors, nurses, advanced practitioners and auxiliary staff. Interviews were recorded, transcribed and analysed using inductive thematic analysis. Six themes were identified. Tracheostomy was seen as a skilled, high‐risk procedure best learned through supervised practice. Teamwork, equipment and the environment were considered vital to safety. Staff reported varied approaches to planning and performing the procedure, alongside strategies to prevent and manage complications. Participants reflected on how a guidance device might improve accuracy, drawing on experiences of the benefits and challenges of introducing new technologies into complex clinical settings. This study provides insights into tracheostomy practices, safety strategies and the potential role of guidance devices. Our study also presents a comprehensive end‐user co‐development model for future medical device design
‘Drawing the Line’ : Market Rule and Liberal Planning in the Governance of Capitalism
No full textThis article critically engages with the ‘death of neoliberalism’ debate. Its purpose is not to determine whether neoliberalism is dying or not but to highlight the limitations of the question, ‘is neoliberalism dead?’ The question invites answers that cannot satisfactorily account for both continuities and transformations in capitalist governance. Existing sides of the debate miss the fact that market discipline is a shared commitment of all iterations of liberal governance, not only neoliberalism. Liberal governance, I argue, is a balancing act between market discipline and planning measures to manage the social costs of accumulation. The argument develops in two steps. First, it presents a brief intellectual history of social liberalism – the most explicitly pro-planning voice in the liberal canon – highlighting its similarities with neoliberalism. Second, it theoretically engages with the Open and Political Marxist traditions to discern the drivers behind the state’s concurrent impulse to expand and self-limit the scope of planning
The space of actions, partition metric and combinatorial rigidity
Get PDFWe introduce a natural pseudometric on the space of actions of d-generated groups. In this pseudometric, the zero classes correspond to the weak equivalence classes defined by Kechris, and the metric identification is compact. We achieve this by employing symbolic dynamics and an ultraproduct construction which also facilitates the extension of our results to unitary representations. As a byproduct, we show that the weak equivalence class of every free non-amenable action contains an action that satisfies the measurable von Neumann problem
Deeper Caribbean reef fish communities show greater taxonomic and functional change in dominance structure over a nine-year period
No full textFish communities at greater depths on a reef are thought to be less affected by disturbances that more strongly impact shallower areas. As a result, these deeper communities might be expected to show less change in their diversity and composition over time compared to those in shallow water. To test this hypothesis, we analysed changes in reef fish composition at 5-15 m and 25-40 m on reefs around Utila, Honduras, across two time periods: 2014-2015 and 2022-2023. We estimated taxonomic and functional α- and β- diversity using coverage-based standardisation and Hill-Chao numbers at orders q = 0 (species richness) and q = 2 (inverse Simpson index). Results showed that the α-diversity of fish communities was more consistent at 25-40 m than at shallower depths between the two time periods. However, β-diversity of dominant species and traits (q = 2) increased at greater depths, indicating that deeper fish communities became more distinct from one another in both structure and function, as well as more different from shallower communities at the same sites. Changes in diversity also varied between sites, highlighting the role of sitespecific conditions in shaping and maintaining fish communities across depths. Overall, the findings are not consistent with the expectation that greater depth reduces temporal community variability, and they raise questions about whether depth alone can serve as a refuge for reef fish
Class-imbalanced flow meter fault diagnosis under small samples using reinforcement learning based Mahalanobis Taguchi system
Get PDFFlow meter is one of the most essential sensors in industrial development, energy measurement and environmental protection. Monitoring of flow meter performance can help detect anomalies early and enable timely corrective actions for critical industrial equipment in harsh operating environments. However, flow meter diagnostic models are often prone to overfitting and low accuracy caused by class-imbalanced small-sample data. To address these problems, a reinforcement learning Mahalanobis Taguchi system (RLMTS) model is proposed in this paper, which primarily consists of three modules, namely Mahalanobis space (MS) construction, threshold determination, and sample classification. In the MS module, an initial MS is constructed by selecting variables through orthogonal array design and signal-to-noise ratio analysis. Reinforcement learning is then introduced to adaptively refine the MS which is verified by the Mahalanobis distance. In the threshold determination module, a neural network algorithm is proposed to replace the traditional quality loss function for optimal threshold determination. In the sample classification module, the fault diagnosis of unknown samples is performed using the valid MS and calculated Mahalanobis distance. Experimental results show that the proposed RLMTS is not only suitable for flow meter fault diagnosis under different class-imbalance ratios with different small sample sizes, but also demonstrates a better diagnostic performance, stronger robustness, and broader applicability compared to the 19 benchmark diagnosis models. The use of RLMTS therefore guarantees stable operation of the flow meters, contributing to energy savings and environmental protection
On the performance of large language models on introductory programming assignments
No full textRecent advances in artificial intelligence (AI), machine learning (ML), and natural language processing (NLP) have led to the development of a new generation of Large Language Models (LLMs) trained on massive amounts of data. Commercial applications (e.g., ChatGPT) have made this available to the general public, enabling the use of LLMs to produce high-quality texts for academic and professional purposes. Educational institutions are increasingly aware of students’ use of AI-generated content and are researching its impact and potential misuse. Computer Science (CS) and related fields are particularly affected, as LLMs can also generate programming code in various languages. To understand the potential impact of publicly available LLMs in CS education, we extend our previously introduced CSEPrompts (Raihan et al. 2024), a framework comprising hundreds of programming exercise prompts and multiple-choice questions from introductory CS and programming courses. We provide experimental results on CSEPrompts, evaluating the performance of several LLMs in generating Python code and answering basic computer science and programming questions, offering insights into the implications of this technology for CS education
Elsa and the Wolfman
No full textA book of poems examining and celebrating the father-daughter relationshi
Predicting shoreline changes using deep learning techniques with Bayesian optimisation
No full textAccurate prediction of shoreline change is vital for effective coastal planning and management, especially under increasing climate variabilities. This study explores the applicability of deep learning (DL) techniques, particularly Long Short-Term Memory (LSTM) and Convolutional Neural Network-LSTM (CNN-LSTM) models, for shoreline forecasting at monthly to inter-annual timescales, under two modelling approaches—direct input (DI) and autoregressive (AR). All models demonstrated the ability to reproduce temporal shoreline variability, while the autoregressive DL models were performing better. Further, a noise impact assessment revealed that seasonal decomposition and noise filtering significantly enhanced the model performance. In particular, the models using 52-week data decomposition and residual noise reduction improved the model performance. The reduction of data noises also resulted in narrower ensemble prediction envelopes, indicating that ensemble candidate models behave with low diversity. The temporal data resolution analysis showed that lower data resolutions reduce the predictive performance of the model and at least fortnightly data are required to satisfactorily capture the trend of variability of the shoreline position at this beach. The use of ensemble predictions, derived from a selected subset of model trials based on their collective performance, proved beneficial by capturing diverse temporal behaviours, thereby offering a quasi-probabilistic forecast with minimal computational cost. Overall, the study underscores the potential of DL models, particularly with autoregressive architectures, for reliable and transferable shoreline change prediction. It also emphasizes the importance of data quality, resolution, and preprocessing in improving model robustness, laying the groundwork for future research into use of DL in multi-scale shoreline predictions
What works? A grounded theory investigation of training non‐psychology staff in using Solution‐Focused Brief Therapy
No full textObjective: As part of a task‐sharing strategy, clinical psychologists are becoming increasingly expected to offer therapy training for staff in health care services to develop psychological mindedness to increase access and provision of psychological support for clients. The current study explored how 10 staff working in health care settings experienced Solution‐Focused Brief Therapy (SFBT) training and how they subsequently used it. Methods: One‐to‐one semi‐structured interviews were conducted with 10 participants; a constructivist grounded theory (GT) approach was used to generate a model based on the participants' reflections after the training. Results: Staff shared how they felt they needed evidence of SFBT effectiveness in order to believe that learning a new model would be worth the required investment. They also found realistic role modelling that was relevant to their context to be particularly convincing as well as regular support from their peers and multidisciplinary meetings. Participants also shared some barriers to using SFBT in practice, including time‐restricted clinics, service pressures and challenging clients. Conclusion: The model describes a complex dynamic between personal, interpersonal and systemic factors that influenced the staff members' individual decision to abandon the more familiar medical model that represented a sense of comfort and safety. The study includes recommendations for how clinical psychologists can address the identified facilitators and barriers to facilitate more effective training programmes and training transfer