26207 research outputs found
Sort by
Universal bovine identification via depth data and deep metric learning
No full textThis paper proposes and evaluates, for the first time, a top-down (dorsal view), depth-only deep learning system for accurately identifying individual cattle and provides associated code, datasets, and training weights for immediate reproducibility. An increase in herd size skews the cow-to-human ratio at the farm and makes the manual monitoring of individuals more challenging. Therefore, real-time cattle identification is essential for the farms and a crucial step towards precision livestock farming. Underpinned by our previous work, this paper introduces a deep-metric learning method for cattle identification using depth data as a novel biometric measure acquired with an off-the-shelf 3D camera. In contrast to our previous work, which was limited to breeds with distinct coat patterns, this study introduces a breed-agnostic pipeline for universal cattle identification. The results show that depth, as a biometric, can potentially broaden the real-world applicability of our method to the rest 68% of UK cattle breeds that lack distinctive coat patterns. The method relies on Convolutional Neural Network (CNN) and Multi-Layered Perceptron (MLP) backbones that learn well-generalised embedding spaces from the body shape to differentiate individuals — requiring neither species-specific coat patterns nor close-up muzzle prints for operation. The network embeddings are clustered using a simple algorithm such as -Nearest Neighbours (-NN) for highly accurate identification, thus eliminating the need to retrain the network for enrolling new individuals. We evaluate two backbone architectures, Residual Neural Network (ResNet), as previously used to identify Holstein Friesians using RGB images, and PointNet, which is specialised to operate on 3D point clouds. We also present CowDepth2023, a new dataset containing 21,490 synchronised colour-depth image pairs of 99 cows, to evaluate the backbones. Both ResNet and PointNet architectures, which consume depth maps and point clouds, respectively, led to high accuracy that is on par with the coat pattern-based backbone. This new universal methodology also addresses the case of all-black and all-white breeds, where the previous coat pattern-based approach fell short. The ResNet colour backbone resulted in 99.97% -NN identification accuracy, while the PointNet accuracy was 99.36%. Furthermore, we also show that the PointNet architecture is robust to noise and missing data by significantly reducing the number of 3D points and observing the drop in accuracy. Our research indicates that these techniques can identify animals using dorsal-view depth maps alone. Regardless of the substantial inter-class variety in the body shape, we show that the models spatially rely on similar body surfaces using Gradient-weighted Class Activation Mapping (Grad-CAM) and Point Cloud Saliency Mapping (PC-SM)
The attitudes of midwives towards NHS fee-charging and data-sharing policies for migrant mothers in the UK: A Q-methodology study
No full textIn 2012, the UK government announced its intentions to create a “hostile environment” for its undocumented residents. As part of this, fee-charging and data-sharing practices were introduced within maternity care for individuals not “ordinarily resident” within the state. Despite efforts to separate the provision of care from the identification of chargeable persons, evidence indicates that midwives are often responsible for implementing these policies. This has transformed the attitudes of midwives into key health determinants for migrant women, yet they remain under-researched. In response to this knowledge deficit, a Q-methodological approach was taken to capture and analyze midwives' attitudes towards fee-charging and data-sharing practices. Twenty-one midwives were purposively recruited and asked to rank a range of statements relating to these policies and their wider impact, based on level of agreement. These statement rankings were then factor analyzed and four distinct belief clusters were identified: Dismayed Policy Sceptics, Medical Tourism Critics, NHS Value Preservers, and Citizen Partisans. Although some areas of consensus were evident, the interpretation of these factors highlighted significant variation in midwives’ experiences of fee-charging policies in daily practice. Contrary to previous literature, the Medical Tourism Critics reported feeling grateful that such policies existed, as they protected the NHS from exploitation. Though this constituted a minority opinion, it indicated the need to improve current practices, rather than overhaul existing policies. Reflecting on these findings, this paper presents recommendations to improve efficiency and alleviate conflict around the implementation of fee-charging policies within maternity care
Using polarization to estimate surface normals at air–water interfaces for correction of refraction in seafloor imaging
No full textThe retrieval of sea surface normal vectors using shape-from-polarization is investigated for the purpose of correcting for refraction at the water–air interface when imaging from above the water. In shallow clear water and overcast conditions, spectral longpass filtering (using a hard-coated 850 nm cut-on wavelength filter) is demonstrated to 1) avoid artifacts from the ground in the measured polarization state, and 2) reduce polarization from water-leaving radiance sufficiently to derive shape information exclusively from the polarization produced by specular reflection. The dependence of the method on meteorological conditions is studied. Measurements are performed with a commercial polarization filter array (PFA) camera. Due to the decreasing PFA efficiency towards the near-infrared, rigorous characterization and calibration measurements were performed and recommendations (e.g., on the f-number) elaborated. Overcoming the paraxial approximation, normal vectors are then retrieved with systematic errors of 0.1∘ (image center) to 0.5∘−0.8∘ (edges/corners) for a flat water surface. An image of the sea floor corrected for surface refraction shows maximum displacements of 10–20 pixels only (corresponding to 0.25∘) with respect to a validation image without water
Using neuroevolution for designing soft medical devices
No full textSoft robots can exhibit better performance in specific tasks compared to conventional robots, particularly in healthcare related tasks. However, the field of soft robotics is still young, and designing them often involves mimicking natural organisms or relying heavily on human experts’ creativity. A formal automated design process is required. The use of neuroevolution-based algorithms to automatically design initial sketches of soft actuators that can enable the movement of future medical devices, such as drug-delivering catheters, is proposed. The actuator morphologies discovered by algorithms like Age-Fitness Pareto Optimisation, NeuroEvolution of Augmenting Topologies (NEAT), and Hypercube-based NEAT (HyperNEAT) were compared based on the maximum displacement reached and their robustness against various control methods. Analyzing the results granted the insight that neuroevolution-based algorithms produce better-performing and more robust actuators under diverse control methods. Specifically, the best-performing morphologies were discovered by the NEAT algorithm
An LSTM approach to deciphering irrigation operations from remote sensing and groundwater levels records
No full textAgricultural irrigation, the largest consumptive water user, significantly impacts terrestrial energy and water cycle, atmospheric boundary layer and the sustainability of water resources management. However, irrigation records usually lack the necessary detail in terms of amount, location, time and source with adequate spatial and temporal resolution that are required for understanding farmers’ irrigation behavior and representing irrigation in hydrologic models. This study addresses the irrigation scheduling gap by leveraging in situ groundwater level records of index wells and multi-source remote sensing observations. We used a Bi-directional Long Short-Term Memory (LSTM) network to capture the temporal relationship between groundwater fluctuations and land surface responses to irrigation. We trained the LSTM model to detect irrigation events based on groundwater level changes in the High Plains region of Nebraska and Kansas from 2001 to 2020. Using Integrated Gradients, an Explainable AI (XAI) technique, we identified that precipitation, MODIS evapotranspiration (ET), and Near-Infrared NIR reflectance are critical factors in detecting irrigation, with antecedent rainfall reducing irrigation likelihood. This framework enables allocation of long-term irrigation amounts to individual events, allows hydrologic models to assimilate irrigation dataset to assess irrigation impacts, and improves irrigation behavior representation in water resources management
Temporary recurring closures and changing mobility patterns: A quasi-experimental study of the impacts of London’s Covid-19 school streets on travel to school
No full textWith over 500 schemes installed since March 2020, School Streets have been one of the most significant street experiments conducted in London. As temporary and recurring closures to streets these schemes provide a novel typology of street experiment which may be applicable to other contexts. In research on urban street experiments, questions have been raised about the extent to which such schemes can contribute to wider mobility transitions. Through a quasi-experimental analysis of school travel data, this study seeks to assess this question in relation to London’s School Streets schemes. It asks to what extent have these schemes reduced the use of private motor vehicles and increased the uptake of active modes of travel to school. The analysis finds positive but modest results on this count, with some evidence that School Streets have helped to prevent a ‘car-based recovery’ from Covid-19 in London. It goes on to reflect on the implications of this for the wider study of the impacts of street experiments on urban mobility
Learning to collaborate, preparing students for practice in multi-disciplinary teams
No full textUWE’s College of Arts, Technology and Environment offers awards related to most construction disciplines, enabling interdisciplinary collaborative working that reflects industry. In Collaborative Practice students from disciplines relating to the construction industry work collaboratively, to challenge concerns highlighted in various industry reports. Guided by construction professionals, students develop and apply a range of interdisciplinary skills. The academic aim of the module develops an understanding of the roles and responsibilities of respective members of the construction team and their interactions through different stages of projects, provided through a mixture of lectures and group workshop activities. The assessment comprises a group presentation in mixed-discipline groups and a collection of reflective writing on academic learning and collaborative working. This module makes a concerted effort to include everyone. Student background is diverse in not only discipline, but in ethnicity and experience. The module team is aware of the distrust that might exist with preconceived ideas about ‘others,’ whether other professionals or other backgrounds. A soft module aim is to address those notions directly and to encourage healthy, collaborative working across the team. This paper will explore how the module enables collaboration and reflexive practice, and as a result, builds student confidence, promotes inclusion and provides preparation for working in multi-disciplinary teams
An investigation into the experiences of those paramedics rotating in primary care from South Western Ambulance Service: A qualitative study
Get PDFIntroduction: The Additional Roles Reimbursement Scheme expands roles, including that of paramedics, and funding into the existing workforce in primary care. This has been laid out in the five-year general practice contract reform framework, with the goal of transforming and evolving the way in which primary care is delivered. Paramedics are rotating from the ambulance service into primary care to help tackle workforce shortages. The aim of this qualitative study was to investigate the experience of those paramedics rotating into primary care from the ambulance service. Methods: This qualitative study utilised convenience sampling of paramedics who were on rotation in primary care within one ambulance service. Eight semi-structured interviews took place. Results: Three key themes of supervision, education and workforce planning were established. Day-to-day supervision was often seen; however, more formal supervision, such as having a designated mentor and completing the first-contact practitioner (FCP) portfolio, was inconsistent. There were clear core skill educational gaps between ambulance paramedics and those that work in primary care. A workforce model, and how this affects the wider system, was discussed, including issues of retention, decision making and referrals. Conclusion: Inconsistent supervision in primary care for FCP roles is evident across disciplines, with physiotherapists acknowledging the same shortcomings. There is a need for more structured support, with access to a mentor / supervision with any FCP role. Within the primary care training period there is a need for a training needs analysis and educational days to support core skills gaps. Due to the positive workforce planning, it is seen that rotations in primary care help to retain staff and have some clear system benefits. To further this, an expansion of the rotations into other areas within the NHS should be considered
Application of artificial intelligence for resilient and sustainable healthcare system: Systematic literature review and future research directions
Get PDFRecent years have witnessed increased pressure across the global healthcare system during the COVID-19 pandemic. The COVID-19 pandemic shattered existing healthcare operations and taught us the importance of a resilient and sustainable healthcare system. Digitisation, specifically adoption of Artificial Intelligence (AI) has positively contributed to developing a resilient healthcare system in recent past. To understand how AI contributes to building a resilient and sustainable healthcare system, this study based on systematic literature review of 89 articles extracted from Scopus and Web of Science databases is conducted. The study is organised around several key themes such as applications, benefits, and challenges of using AI technology in healthcare sector. It is observed that AI has wide applications in radiology, surgery, medical, research, and development of healthcare sector. Based on the analysis, a research framework is proposed using an extended Antecedents, Practices, and Outcomes (APO) framework. This framework comprises AI applications’ antecedents, practices, and outcomes for building a resilient and sustainable healthcare system. Consequently, three propositions are drawn in this study. Furthermore, our study has adopted the theory,contextand methodology (TCM) framework to provide future research directions, which can be used as a reference point for future studies
Exploring exudate absorption via sessile droplet dynamics in porous wound dressings
Get PDFChronic wounds, typically defined as those that fail to reduce in size by at least 40% within a month, present a significant global socioeconomic challenge. In clinical practice, it is widely recognized that maintaining an optimal moisture balance in the wound while managing excess exudate is crucial for wound healing. Therefore, the selection of wound dressings is a key tool in wound management, which is based on their ability to sustain this delicate equilibrium. However, there is a notable lack of fundamental studies on the interaction between wound exudate and dressings, which limits the availability of evidence-based guidance for clinical practitioners. Thus, the present investigation explores how wound exudate interacts with different commercially available wound dressings to optimize wound management through a deep understanding of exudate-air interface dynamics in contact with the dressing material. Employing high-resolution imaging, the research delves into the behaviour of quasi-sessile droplets on various porous materials, analysing the impacts of exudate viscosity, blood sugar levels, and exudate volume. The findings reveal that droplet absorption rates depend on exudate properties and dressing materials. Notably, cellulose-based dressings outperform alginate and polyester-based alternatives in terms of wettability and imbibition capacity, with a performance improvement of at least 48%. Furthermore, increased exudate viscosity and elevated blood sugar are associated with longer absorption times, with increases of 51% and 38%, respectively. The study also identifies that absorption completion time increases exponentially with fibre diameter but decreases with greater pore radius and higher porosity. The overall findings can aid clinicians with quantitative insights to optimize the selection of wound dressings, thereby enhancing the healing of chronic wounds