114612 research outputs found
Sort by
Clustered federated learning in Internet of Things : convergence analysis and resource pptimization
Get PDFFederated learning (FL) framework enables user devices to collaboratively train a global model based on their local datasets without privacy leak. However, the training performance of FL is degraded when the data distributions of different devices are incongruent. Fueled by this issue, we consider a clustered FL (CFL) method where the devices are divided into several clusters according to their data distributions and are trained simultaneously. Convergence analysis is conducted, which shows that the clustered model performance depends on cosine similarity, device number per cluster, and device participation probability. Besides, to quantify the training performance, the utility of clustered model training is defined based on the analysis results. Then, aiming at optimizing the system utility, a joint problem of resource allocation and device clustering is formulated, which is solved by decoupling it into two sub-problems. First, given the results of device clustering, a low-complexity iterative algorithm based on the convex optimization theory is proposed to make the bandwidth allocation and the transmit power control. Then, according to the individual stability, a coalition formation algorithm is proposed for the device clustering. Finally, the real-data experiments on the classification tasks (e.g. MNIST, CIFAR-10, CIFAR-100) validate the results of convergence analysis and advantages of the proposed algorithm in terms of the test accuracy
Estimation of auto-covariance of log hydraulic conductivity from Generalized Sub-Gaussian porosity and particle size random fields
Get PDFWe derive analytical formulations relating the spatial covariance (CY) of (log-transformed) hydraulic conductivities to auto- and cross-covariances of porosity (ϕ) and representative soil particle sizes within the framework of the classical Terzaghi model. The latter provides an empirical relationship which is widely used to obtain conductivity estimates. We frame the study within recent stochastic approaches and conceptualize appropriate transformations of ϕ and representative soil particle size as Generalized Sub-Gaussian (GSG) spatially cross-correlated random processes. Consistency of the theoretical framework against sample distributions of ϕ and particle size is assessed through the analysis of field data. A perturbation-based approach yields workable expressions of CY upon truncating the otherwise exact analytical solution at given orders of approximations. Our analytical (truncated) log-conductivity covariance is in agreement with its Monte Carlo-based counterpart. A Global Sensitivity Analysis relying on classical Sobol indices quantifies the relative importance of all parameters embedded in the formulation of CY. We show that parameters driving the GSG nature of the distribution of (transformed) porosity are key to the main features of CY. We also document the relevance of properly capturing emergences of possible cross-correlations between ϕ and representative particle size to reconstruct conductivity fields
Exploring the experiences and preferences of South Asian patients' of primary care in England since COVID‐19
Get PDFIntroduction: Remote (digital and/or telephone) access and consultation models are being driven by national policy with the goal being that the National Health Service operate on a remote‐first (digital‐first) basis by 2029. Previous research has suggested that remote methods of access to care and consulting may act to widen health inequalities for certain patients and/or groups such as those from ethnic minorities. South Asian (SA) patients comprise the largest ethnic minority group in England. Understanding the experiences and needs of this group is critical to ensuring that general practice can deliver equitable, quality health care. Methods: Qualitative study. 37 participants (from Indian, Pakistani and/or Bangladeshi background) were recruited to take part in either in‐person preferred language focus groups or remote semistructured interviews in the English language. Thematic analysis was conducted to identify themes in the qualitative data. Findings: Three major interlinked themes were identified: (1) reduced access, (2) reduced patient choice and (3) quality and safety concerns. The findings highlight access issues split by (i) general issues with appointment access via any remote means and (ii) specific issues related to language barriers creating additional barriers to access and care. Some patients valued the convenience of remote access but also raised concerns regarding appointment availability and reduced patient choice. Face‐to‐face consultations were preferable but less available. The findings underscore how participants perceived remote care to be of lesser quality and less safe. Concerns were greatest for those with limited English proficiency (LEP), with the removal of non‐verbal aspects of communication and ‘hands‐on’ care leading to perceptions of reduced psycho‐social safety. Conclusion: SA patients' experiences of remote‐led primary care access and care delivery were negative with only a minority viewing it positively and for certain limited scenarios. Face‐to‐face models of care remain the preferred mode of consultation, particularly for those with LEP. Hybrid models of access offer patients the greatest choice, and are likely to meet the varying needs of the South‐Asian patient population going forwards. The remote first approach to primary care may be achievable as a service ideal, but its limitations need to be recognised and accounted for to ensure that primary care can be an equitable service, both now and in the future. Public Contribution: Members of the public were involved in all phases of research in the study. This included co‐working in partnership throughout the study including, reviewing patient‐facing documents, recruiting participants, data facilitation, translation work, interpretation of the data and co‐authors on this manuscript. The key to the success of our study was collaborative teamwork, which involved experienced members of the public with SA cultural knowledge working together with and integral to the research team for all components
Zeta functions in higher Teichmüller theory
Get PDFIn this note we introduce zeta functions and L-functions for discrete and faithful representations of surface groups in PSL(d, R), for d≥3. These are natural generalizations of the well known classical Selberg zeta function and L-function for Fuchsian groups, corresponding to the case d=2. We show that these complex functions have meromorphic extensions to the entire complex plane C
Fetal loss and long-term maternal morbidity and mortality : a systematic review and meta-analysis
Get PDFBackground:
Evidence suggests common pathways between pregnancy losses and subsequent long-term maternal morbidity, rendering pregnancy complications an early chronic disease marker. There is a plethora of studies exploring associations between miscarriage and stillbirth with long-term adverse maternal health; however, these data are inconclusive.
Methods and findings:
We systematically searched MEDLINE, EMBASE, AMED, BNI, CINAHL, and the Cochrane Library with relevant keywords and MeSH terms from inception to June 2023 (no language restrictions). We included studies exploring associations between stillbirth or miscarriage and incidence of cardiovascular, malignancy, mental health, other morbidities, and all-cause mortality in women without previous pregnancy loss. Studies reporting short-term morbidity (within a year of loss), case reports, letters, and animal studies were excluded. Study selection and data extraction were performed by 2 independent reviewers. Risk of bias was assessed using the Newcastle Ottawa Scale (NOS) and publication bias with funnel plots. Subgroup analysis explored the effect of recurrent losses on adverse outcomes. Statistical analysis was performed using an inverse variance random effects model and results are reported as risk ratios (RRs) with 95% confidence intervals (CIs) and prediction intervals (PIs) by combining the most adjusted RR, odds ratios (ORs) and hazard ratios (HRs) under the rare outcome assumption. We included 56 observational studies, including 45 in meta-analysis. There were 1,119,815 women who experienced pregnancy loss of whom 951,258 had a miscarriage and 168,557 stillbirth, compared with 11,965,574 women without previous loss. Women with a history of stillbirth had a greater risk of ischaemic heart disease (IHD) RR 1.56, 95% CI [1.30, 1.88]; p < 0.001, 95% PI [0.49 to 5.15]), cerebrovascular (RR 1.71, 95% CI [1.44, 2.03], p < 0.001, 95% PI [1.92, 2.42]), and any circulatory/cardiovascular disease (RR 1.86, 95% CI [1.01, 3.45], p = 0.05, 95% PI [0.74, 4.10]) compared with women without pregnancy loss. There was no evidence of increased risk of cardiovascular disease (IHD: RR 1.11, 95% CI [0.98, 1.27], 95% PI [0.46, 2.76] or cerebrovascular: RR 1.01, 95% CI [0.85, 1.21]) in women experiencing a miscarriage. Only women with a previous stillbirth were more likely to develop type 2 diabetes mellitus (T2DM) (RR: 1.16, 95% CI [1.07 to 2.26]; p < 0.001, 95% PI [1.05, 1.35]). Women with a stillbirth history had an increased risk of developing renal morbidities (RR 1.97, 95% CI [1.51, 2.57], p < 0.001, 95% [1.06, 4.72]) compared with controls. Women with a history of stillbirth had lower risk of breast cancer (RR: 0.80, 95% CI [0.67, 0.96], p-0.02, 95% PI [0.72, 0.93]). There was no evidence of altered risk of other malignancies in women experiencing pregnancy loss compared to controls. There was no evidence of long-term mental illness risk in women with previous pregnancy losses (stillbirth: RR 1.90, 95% CI [0.93, 3.88], 95% PI [0.34, 9.51], miscarriage: RR 1.78, 95% CI [0.88, 3.63], 95% PI [1.13, 4.16]). The main limitations include the potential for confounding due to use of aggregated data with variable degrees of adjustment.
Conclusions:
Our results suggest that women with a history of stillbirth have a greater risk of future cardiovascular disease, T2DM, and renal morbidities. Women experiencing miscarriages, single or multiple, do not seem to have an altered risk
Inspection techniques using scanning acoustic microscopy for silver sintering applications in power electronic modules
Get PDFThis work presents a practical perspective to Scanning Acoustic Microscopy (SAM) for Ag-sintering applications in power electronic modules and devices. It aims to present typical challenges to consider whilst performing C-SAM for inspecting and analysing microstructures within power electronic devices, with emphasis on silver-sintered interconnects. The study presents a vast number of acoustic images (discrete devices in TO-247 packages, power modules, and also custom samples of substrates and devices), where several defects are identified, including voids, cracks, and delamination. Challenges for Ag-sintered layers inspection related to advanced surface finishes are also revealed and discussed in detail. The suitability of C-SAM for inspecting small and large areas of Ag-sintered interfaces is discussed, and showcased with custom samples, and a commercial power module
The Benchmarking Exercise Programme for Older People (BEPOP) : design, results and recommendations from the first wave of data collection
No full textObjectives:
The Benchmarking Exercise Programme for Older People (BEPOP) service improvement project seeks to determine and promote the exercise training characteristics associated with positive outcomes for resistance exercise for older people living with, or at risk of, sarcopenia or physical frailty.
Methods:
Mixed-methods service improvement project. Ten UK National Health Service physiotherapist-led therapy services delivering exercise interventions for older people submitted anonymized data for up to 20 consecutive patients. A multidisciplinary expert panel generated a report and recommendations with site-specific benchmarking data and feedback. In parallel, participating physiotherapy team members were interviewed to elicit feedback on BEPOP rationale, processes and perceived value.
Results:
Data from 188 patients were included, mean age 80 years (range 60-101). 115 (61%) received objective assessment of strength-based physical performance. Bodyweight exercises (173 [92%]) and resistance bands (49 [26%]) were the commonest exercise modalities. Exercises progressed predominantly through increased repetitions (163 [87%]) rather than increased load. 50 (30%) had no reassessment of outcomes; only 68 (41%) were signposted to follow-on exercise services. Staff interviews identified themes around knowledge, diagnosis, data collection and practice reflection.
Conclusion:
BEPOP was feasible to deliver and generated actionable insights for service improvement via improved diagnosis, measurement and progression of resistance exercise
A shorter version of the revised Francis Psychological Type and Emotional Temperament Scales (FPTETS-r)
Get PDFThe Francis Psychological Type and Emotional Temperament Scales (FPTETS) operationalize the psychological-type model of personality alongside emotional temperament. The scales have been widely used in research as continuous variables that explain a wide range of religious beliefs and attitudes. The full instrument consists of five 10-item scales so a shorter version would be useful in longer surveys where completion time needs to be minimized. This study uses data from 700 Church of England clergy who completed the revised version of the FPTETS to reduce the 10-item scales to 6-item scales. Ant colony optimization was found to be a better way of selecting the final items than reliability optimization alone because it balanced individual scale reliabilities with maintaining the factor structure of the overall instrument. The selected scales were validated using data from 1,194 lay people from the Church of England, and two samples of 884 clergy and 2,765 lay people from the Episcopal Church (USA). The short scales are commended for use where the need is for continuous scale scores rather than producing psychological typologies
A novel hybrid CNN methodology for automated leaf disease detection and classification
No full textPlant leaf diseases are challenging to categorize due to the complexity of the pattern variations and the high degrees of inter‐class similarity. Plant ailments harm food quality and production. To ensure the quality and quantity of harvests, it is essential to protect plants from disease. Detection of diseases at an early stage is the main and the most complex task for farmers due to common morphological properties like colour, shape, texture, and edges. In this study, a Hybrid Deep Learning model named Hybrid‐Convolutional Support Machine (H‐CSM) based on ‘Support Vector Machine (SVM)’, ‘Convolutional Neural Network (CNN)’ and ‘Convolutional Block Attention Module (CBAM)’ is proposed for the early diagnosis and classification of leaf diseases in plants leaf. The suggested model can initially identify different plant leaf illnesses, although it is not constrained to these. A database of pictures of plant leaves is used to test the suggested method based on different evaluation parameters. The results were highly promising, with an accuracy of up to 98.72% which has been increased by applying better learning methods. Farmers can quickly identify 36 common diseases with a little instruction for 14 plant categories, enabling them to take prompt preventive measures using the proposed method
Clinical outcomes and adverse events of bariatric surgery in adults with severe obesity in Scotland : the SCOTS observational cohort study
Get PDFBackground Bariatric surgery is a common procedure worldwide for the treatment of severe obesity and associated comorbid conditions but there is a lack of evidence as to medium-term safety and effectiveness outcomes in a United Kingdom setting. Objective To establish the clinical outcomes and adverse events of different bariatric surgical procedures, their impact on quality of life and the effect on comorbidities. Design Prospective observational cohort study. Setting National Health Service secondary care and private practice in Scotland, United Kingdom. Participants Adults (age >16 years) undergoing their first bariatric surgery procedure. Main outcome measures Change in weight, hospital length of stay, readmission and reoperation rate, mortality, diabetes outcomes (HbA1c, medications), quality of life, anxiety, depression. Data sources Patient-reported outcome measures, hospital records, national electronic health records (Scottish Morbidity Record 01, Scottish Care Information Diabetes, National Records Scotland, Prescription Information System). Results Between December 2013 and February 2017, 548 eligible patients were approached and 445 participants were enrolled in the study. Of those, 335 had bariatric surgery and 1 withdrew from the study. Mean age was 46.0 (9.2) years, 74.7% were female and the median body mass index was 46.4 (42.4; 52.0) kg/m2. Weight was available for 128 participants at 3 years: mean change was −19.0% (±14.1) from the operation and −24.2% (±12.8) from the start of the preoperative weight-management programme. One hundred and thirty-nine (41.4%) participants were readmitted to hospital in the same or subsequent 35 months post surgery, 18 (5.4% of the operated cohort) had a reoperation or procedure considered to be related to bariatric surgery gastrointestinal complications or revisions. Fewer than five participants (<2%) died during follow-up. HbA1c was available for 93/182 and diabetes medications for 139/182 participants who had type 2 diabetes prior to surgery; HbA1c mean change was −5.72 (±16.71) (p = 0.001) mmol/mol and 65.5% required no diabetes medications (p < 0.001) at 3 years post surgery. Physical quality of life, available for 101/335 participants, improved in the 3 years post surgery, mean change in Rand 12-item Short Form Survey physical component score 8.32 (±8.95) (p < 0.001); however, there was no change in the prevalence of anxiety or depression. Limitations Due to low numbers of bariatric surgery procedures in Scotland, recruitment was stopped before achieving the intended 2000 participants and follow-up was reduced from 10 years to 3 years. Conclusions Bariatric surgery is a safe and effective treatment for obesity. Patients in Scotland, UK, appear to be older and have higher body mass than international comparators, which may be due to the small number of procedures performed. Future work Intervention studies are required to identify the optimal pre- and post surgery pathway to maximise safety and cost-effectiveness. Study registration This study is registered as ISRCTN47072588. Funding details This award was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme (NIHR award ref: 10/42/02) and is published in full in Health Technology Assessment; Vol. 28, No. 7. See the NIHR Funding and Awards website for further award information