Search Results (55)

Background/Objectives: Breast cancer remains a leading cause of cancer-related mortality among women worldwide. This study presents a systematically justified multi-modal breast cancer classification pipeline that combines established, physically motivated preprocessing operations, modality-specific deep learning models, late-fusion inference, and a deployment-aware federated learning evaluation. Rather than introducing new image restoration or federated optimization algorithms, this work formalizes how standard preprocessing methods can be organized according to the dominant degradation characteristics of ultrasound, MRI, and mammography, and evaluates their contribution under centralized and simulated federated learning settings. Methods: Patient-wise stratified five-fold cross-validation was applied across ultrasound (BUSI, $n=780$), dynamic contrast-enhanced MRI (DUKE, $n=922$), and mammography (CBIS-DDSM, $n=400$). A five-algorithm federated learning comparison, including FedAvg, FedProx, SCAFFOLD, FedNova, and FP16-FedAvg, was conducted under IID and non-IID conditions using a Dirichlet distribution with $\alpha =0.5$. The evaluation reports diagnostic performance together with per-round training time, communication time, latency-related measurements, and cumulative bandwidth. Ablation experiments, McNemar’s test, Cohen’s h effect sizes, and confidence intervals were used to support the analysis. Results: Per-modality models achieved 92.50 ± 1.2%, 90.63 ± 1.5%, and 92.00 ± 1.3% accuracy for ultrasound, MRI, and mammography, respectively, with statistically significant improvements over the corresponding baselines according to McNemar’s test ($p<0.05$). Weighted late fusion achieved 93.10 ± 1.1% accuracy and improved performance compared with the best individual modality ($p=0.031$). FP16 transmission reduced cumulative bandwidth from 8.14 GB to 1.23 GB ($-84.9\%$) without a statistically significant performance difference compared with FP32 transmission ($p=0.74$), while SCAFFOLD achieved the highest non-IID accuracy (90.50%). Conclusions: The findings demonstrate internal technical validity and deployment-relevant trade-offs, but they should be interpreted cautiously because the federated evaluation is simulation-based, key-slice extraction may require annotation-assisted assumptions, and external multi-center validation remains necessary before clinical deployment. Reported improvements are statistically significant in several comparisons, but corresponding Cohen’s h effect sizes are small, and clinical meaningfulness requires independent validation rather than inference from p-values alone. Full article

With rising incidence in recent years, Listeriosis, a severe foodborne disease in humans primarily transmitted through ready-to-eat (RTE) foods contaminated with Listeria monocytogenes, became the most severe zoonotic disease in the European Union (EU) in 2024 with the highest hospitalization and mortality rates, prompting stricter regulatory requirements under Regulation (EC) No 2073/2005 and its recent amendments. This systematic literature review aimed to evaluate the availability, validity and quality of published challenge test data on the growth potential and maximum growth rate of Listeria monocytogenes in RTE foods to identify data gaps and, if possible, to support the derivation of a classification of RTE foods into the two existing regulatory categories, a and b (not able and able to support the growth of Listeria monocytogenes). Conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and the Cochrane Handbook, a comprehensive database search was done to identify eligible challenge test studies on Listeria monocytogenes growth in RTE foods, followed by structured screening and quality assessment based on the EURL Lm Technical Guidance Documents. A limited and heterogeneous body of published challenge test data on the growth potential and maximum growth rate of Listeria monocytogenes in RTE foods was identified, with substantial data gaps across multiple food groups, precluding meta-analysis and limiting regulatory applicability under the current regulations. Overall, the available literature is insufficient to reliably support regulatory classification or to enable direct extrapolation by food business operators (FBO), underscoring the need for product-specific investigations and food group-specific guidance for food safety. Full article

Background/Objectives: Early and accurate detection of breast cancer is essential for reducing mortality and improving patient outcomes. However, the manual interpretation of breast ultrasound images is challenging due to image variability, noise, and inter-observer subjectivity. This study aims to address these limitations by developing an automated and interpretable computer-aided diagnosis (CAD) system. Methods: We propose an automated and interpretable computer-aided diagnosis (CAD) system that integrates ensemble transfer learning with Vision Transformer architectures. The system combines the Data-Efficient Image Transformer (Deit) and Vision Transformer (ViT) through concatenation-based feature fusion to exploit their complementary representations. Preprocessing, normalization, and targeted data augmentation enhance robustness, while Gradient-weighted Class Activation Mapping (Grad-CAM) provides visual explanations to support clinical interpretability. The proposed model is benchmarked against state-of-the-art CNNs (VGG16, ResNet50, DenseNet201) and Transformer models (ViT, DeiT, Swin, Beit) using the Breast Ultrasound Images (BUSI) dataset. Results: The ensemble achieved 96.92% accuracy and 97.10% AUC for binary classification, and 94.27% accuracy with 94.81% AUC for three-class classification. External validation on independent datasets demonstrated strong generalizability, with 87.76%/88.07% accuracy/AUC on BrEaST, 86.77%/85.90% on BUS-BRA, and 86.99%/86.99% on BUSI_WHU. Performance decreased for fine-grained BI-RADS classification—76.68%/84.59% accuracy/AUC on BUS-BRA and 68.75%/81.10% on BrEaST—reflecting the inherent complexity and subjectivity of clinical subclassification. Conclusions: The proposed Vision Transformer-based ensemble demonstrates high diagnostic accuracy, strong cross-dataset generalization, and clinically meaningful explainability. These findings highlight its potential as a reliable second-opinion CAD tool for breast cancer diagnosis, particularly in resource-limited clinical environments. Full article

Early diagnosis of breast and skin cancers significantly reduces mortality rates, yet manual segmentation remains challenging due to subjective interpretation, radiologist fatigue, and irregular lesion boundaries. This study presents BAS-SegNet, a novel boundary-aware segmentation framework that addresses these limitations through an enhanced deep learning architecture. The proposed method integrates three key innovations: (1) an enhanced CNN-based architecture with a switchable feature pyramid interface, a tunable ASPP module, and consistent dropout regularization; (2) an edge-aware preprocessing pipeline using Sobel-based edge magnitude maps stacked as additional channels with geometric augmentations; (3) a boundary-aware hybrid loss combining Binary Cross-Entropy, Dice, and Focal losses with auxiliary edge supervision from morphological gradients. Experimental validation on the BUSI breast ultrasound and ISIC skin lesion datasets demonstrates superior performance, achieving Dice scores of 0.814 and 0.935, respectively, with IoU improvements of 16.3–22.4% for breast cancer and 8.8–11.5% for skin cancer compared with existing methods. The framework shows particular effectiveness under challenging ultrasound conditions where lesion boundaries are ambiguous, offering significant potential for automated clinical diagnosis support. Full article

Breast cancer is the most frequently diagnosed malignancy among South African women and remains a leading cause of cancer-related death, yet the role of occupation as an independent predictor of mortality has not been evaluated nationally. In this unmatched case–control study using 2011–2019 mortality data, we compared 13,207 breast cancer deaths with 64,849 non-malignant circulatory disease deaths among women aged 30 years and older, classifying usual occupation into major and sub-groups. A multivariable binary logistic regression adjusting for age, year of death, education, province of death and smoking status was conducted. We observed that compared with elementary occupations, breast cancer mortality was significantly higher during 2011–2015 among legislators, senior officials and managers (aMOR = 1.79, 95% CI: 1.36–2.36), clerks (aMOR = 1.75, 95% CI: 1.46–2.11), professionals (aMOR = 1.62, 95% CI: 1.36–1.94), craft and related trades workers (aMOR = 1.55, 95% CI: 1.18–2.05), technicians and associate professionals (aMOR = 1.54, 95% CI: 1.21–1.96), and service workers, shop and market sales workers (aMOR = 1.33, 95% CI: 1.10–1.62), with similar patterns persisting in 2016–2019 where technicians and associate professionals (aMOR = 1.69, 95% CI: 1.44–1.98), legislators, senior officials and managers (aMOR = 1.59, 95% CI: 1.20–2.10), professionals (aMOR = 1.47, 95% CI: 1.23–1.75), clerks (aMOR = 1.43, 95% CI: 1.24–1.65), and service workers (aMOR = 1.34, 95% CI: 1.12–1.61) again showed elevated odds. The sub-occupation analyses for 2011–2015 identified strikingly high risks among building and related trades workers excluding electricians (aMOR = 8.01, 95% CI: 3.06–20.96), legal, social and cultural professionals (aMOR = 3.32, 95% CI: 2.18–5.04), and business and administration professionals (aMOR = 2.18, 95% CI: 1.60–2.97). The results underscore occupation as an essential determinant of breast cancer mortality, highlighting the need for targeted prevention and screening strategies in workers. Full article

Breast cancer is the second leading cause of cancer-related mortality among women. Early detection enables timely treatment, improving survival outcomes. This paper presents a comparative evaluation of federated learning (FL) frameworks for multiclass breast cancer classification using ultrasound images drawn from three datasets: BUSI, BUS-UCLM, and BCMID, which include 600, 38, and 323 patients, respectively. Five state-of-the-art networks were tested, with MobileNet, ResNet and InceptionNet identified as the most effective for FL deployment. Two aggregation strategies, FedAvg and FedProx, were assessed under varying levels of data heterogeneity in two and three client settings. Results from experiments indicate that the FL models outperformed local and centralized training, bypassing the adverse impacts of data isolation and domain shift. In the two-client federations, FL achieving up to 8% higher accuracy and almost 6% higher macro-F1 scores on average that local and centralized training. FedProx on MobileNet maintained a stable performance in the three-client federation with best average accuracy of 73.31%, and macro-F1 of 67.3% despite stronger heterogeneity. Consequently, these results suggest that the proposed multiclass model has the potential to support clinical workflows by assisting in automated risk stratification. If deployed, such a system could allow radiologists to prioritize high-risk patients more effectively. The findings emphasize the potential of federated learning as a scalable, privacy-preserving infrastructure for collaborative medical imaging and breast cancer diagnosis. Full article

Early and accurate detection of breast cancer is essential for reducing mortality rates and improving clinical outcomes. However, deep learning (DL) models used in healthcare face significant challenges, including concerns about data privacy, domain-specific overfitting, and limited interpretability. To address these issues, we propose BreastSwinFedNetX, a federated learning (FL)-enabled ensemble system that combines four hierarchical variants of the Swin Transformer (Tiny, Small, Base, and Large) with a Random Forest (RF) meta-learner. By utilizing FL, our approach ensures collaborative model training across decentralized and institution-specific datasets while preserving data locality and preventing raw patient data exposure. The model exhibits strong generalization and performs exceptionally well across five benchmark datasets—BreakHis, BUSI, INbreast, CBIS-DDSM, and a Combined dataset—achieving an F1 score of 99.34% on BreakHis, a PR AUC of 98.89% on INbreast, and a Matthews Correlation Coefficient (MCC) of 99.61% on the Combined dataset. To enhance transparency and clinical adoption, we incorporate explainable AI (XAI) through Grad-CAM, which highlights class-discriminative features. Additionally, we deploy the model in a real-time web application that supports uncertainty-aware predictions and clinician interaction and ensures compliance with GDPR and HIPAA through secure federated deployment. Extensive ablation studies and paired statistical analyses further confirm the significance and robustness of each architectural component. By integrating transformer-based architectures, secure collaborative training, and explainable outputs, BreastSwinFedNetX provides a scalable and trustworthy AI solution for real-world breast cancer diagnostics. Full article

Background: Ready-to-eat cereals (RTECs) are a large, heterogeneous category of cereals designed to fit into busy lifestyles with minimal preparation time. Methods: This systematic review evaluated nutrient intake data from seven national surveys. Using PubMed and Science Direct (1 January 2004 until 16 September 2024), we investigated RTECs in relation to their contributions to macro, micronutrient and food group intakes, breakfast/diet quality and effects on health with focus on non-communicable disease (NCD) prevention. The search was restricted to Systematic Reviews (SRs), meta-analyses (MAs), randomised controlled trials (RCTs) and observational studies. Fifty-one publications were obtained. Studies related to health outcomes and NCD risk were graded using an updated Scottish Intercollegiate Guidelines Network approach. Results: Grade A evidence: Based on high-quality MA, SRs, or RCTs, this showed that RTEC consumption was associated with improved nutrient intakes (particularly fibre and micronutrients), reduced cardiovascular disease and mortality. One good-quality Grade A meta-analysis showed that total whole grain intake which included cereals was associated with a reduced risk of total cancer. Grade B evidence: Based largely on observational evidence, this showed that RTEC consumption was associated with reduced risk of overweight and obesity, body mass index and composition improvements and type 2 diabetes risk. For food group intakes, breakfast/diet quality and lipid profiles, more well-designed studies were needed (Grade D evidence). Conclusions: There is consistent evidence that RTECs generally have positive or neutral effects on nutritional status and NCD prevention. Strongest evidence exists for RTEC and micronutrient intakes, reduced risk of cardiovascular diseases (CVDs), body weight regulation, and reduced type 2 diabetes risk. Public health messaging should recognise that RTECs, especially whole-grain, higher-fibre and lower-sugar varieties, may help to reinforce micronutrient intakes and a range of health outcomes. Full article

Marine heatwaves (MHWs) are increasingly affecting tropical seas, causing mass coral bleaching and mortality in the wider Caribbean (WC) and eastern tropical Pacific (ETP). This leads to significant coral loss, reduced biodiversity, and impaired ecological functions. Climate models forecast a troubling future for Latin American coral reefs, but downscaled projections for the WC and ETP remain limited. Understanding regional temperature thresholds that threaten coral reef futures and restoration efforts is critical. Our goals included analyzing historical trends in July–August–September–October (JASO) temperature anomalies and exploring future projections at subregional and country levels. From 1940 to 2023, JASO air and ocean temperature anomalies showed significant increases. Projections indicate that even under optimistic scenario 4.5, temperatures may exceed the +1.5 °C air threshold beyond pre-industrial levels by the 2040s and the +1.0 °C ocean threshold beyond historical annual maximums by the 2030s, resulting in severe coral bleaching and mortality. Business-as-usual scenario 8.5 suggests conditions will become intolerable for coral conservation and restoration by the 2030s, with decadal warming trends largely surpassing historical rates, under unbearable conditions for corals. The immediate development of regional and local adaptive coral reef conservation and restoration plans, along with climate change adaptation and mitigation strategies, is essential to provide time for optimistic scenarios to materialize. Full article

Objective: Electronic health records (EHRs) are becoming increasingly important in both academic research and business applications. Recent studies indicate that predictive tasks, such as heart failure detection, perform better when the geometric structure of EHR data, including the relationships between diagnoses and treatments, is considered. However, many EHRs lack essential structural information. This study aims to improve predictive accuracy in healthcare by constructing a Patient Knowledge Graph Ensemble Framework (PKGNN) to analyse ICU patient cohorts and predict mortality and hospital readmission outcomes. Methods: This study utilises a cohort of 42,671 patients from the MIMIC-IV dataset to build the PKGNN framework, which consists of three main components: (1) medical note extraction, (2) patient graph construction, and (3) prediction tasks. Advanced Natural Language Processing (NLP) models, including Clinical BERT, BioBERT, and BlueBERT, extract and integrate semantic representations from discharge summaries into a patient knowledge graph. This structured representation is then used to enhance predictive tasks. Results: Performance evaluations on the MIMIC-IV dataset indicate that the PKGNN framework outperforms state-of-the-art baseline models in predicting mortality and 30-day hospital readmission. A thorough framework analysis reveals that incorporating patient graph structures improves prediction accuracy. Furthermore, an ensemble model enhances risk prediction performance and identifies crucial clinical indicators. Conclusions: This study highlights the importance of leveraging structured knowledge graphs in EHR analysis to improve predictive modelling for critical healthcare outcomes. The PKGNN framework enhances the accuracy of mortality and readmission predictions by integrating advanced NLP techniques with patient graph structures. This work contributes to the literature by advancing knowledge graph-based EHR analysis strategies, ultimately supporting better clinical decision-making and risk assessment. Full article

Background: Antimicrobial resistance (AMR) is an escalating global health threat, projected to cause over 40 million deaths by 2050. ESKAPE pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter spp.) are major contributors to nosocomial infections and AMR. We evaluated the epidemiology and AMR prevalence of ESKAPE pathogens at the University Hospital in Palermo between January 2018 and July 2023, analyzing factors associated with mortality in patients with positive blood cultures. Methods: Microbiological data from all specimen types were collected using the Business Intelligence system Biwer, excluding duplicates. We assessed the prevalence and trends of ESKAPE isolates and AMR over time. Clinical data from hospital discharge forms were used to evaluate factors associated with mortality in patients with ESKAPE-positive blood cultures. Differences in AMR prevalence between blood and non-blood isolates were examined. Results: A total of 11,607 specimens from 4916 patients were analyzed. Most patients were admitted to Internal Medicine (19.4%), the ICU (13.2%), and General Surgery (9.9%). Additionally, 21.5% of the specimens were collected from ICU-admitted patients. Blood cultures accounted for 14.3% of the specimens, urine for 25.3%, respiratory secretions for 22.1%, and skin and mucosal swabs for 20.9%. The prevalence of all isolates increased progressively, peaking in 2021. The vancomycin-resistant E. faecium prevalence was 19.4%, with a significant upward trend, while oxacillin-resistant S. aureus prevalence was 35.0%, showing a significant decline. A. baumannii exhibited high resistance to all antibiotics tested except for colistin and cefiderocol. Carbapenemase resistance was 55.0% in K. pneumoniae, 20.4% in P. aeruginosa, and 4.6% in Enterobacter spp. P. aeruginosa showed a significant decrease in meropenem resistance. K. pneumoniae and A. baumannii bloodstream infections were linked to higher mortality risk. Full article

Background: Infectious bronchitis virus (IBV) causes a significant illness in birds, making it a leading source of financial loss in the poultry business. The objective of this study was to assess the effectiveness of proliposomes (PLs) containing ivermectin (IVM) against IBV using embryonated chicken eggs (ECEs). Methods: A three-factor, two-level (2³) full factorial design was employed; carrier/lipid phase ratio (A), stearyl glycyrrhetinate amount (B), and phospholipid type (C) were studied as independent variables, while product yield (PY), entrapment efficiency (EE), particle size (PS), polydispersity index (PDI), zeta potential (ZP), and cumulative percentage of drug released after 6 h (Q6h) were characterized. The selected formulations (PL2 and PL6) were subjected to further characterizations, including IVM toxicity and anti-viral activity. Results: The PY% ranged from 88.6 ± 2.19% to 98.8 ± 0.45%, EE% was from 71.8 ± 2.01% to 96.1 ± 0.51%, PS was from 330.1 ± 55.65 nm to 1801.6 ± 45.61 nm, PDI was from 0.205 ± 0.06 to 0.603 ± 0.03, ZP was from −18.2 ± 0.60 mV to −50.1 ± 1.80 mV, and Q6h was from 80.95 ± 1.36% to 88.79 ± 2.03%. IVM-loaded PLs had lower toxicity in ECEs than pure IVM; the mortality rate was substantially reduced in IBV-infected ECEs injected with PL2 rather than pure IVM. As further evidence of IVM’s anti-viral action against IBV, quantitative real-time polymerase chain reaction (qRT-PCR) revealed that the PL2-treated group exhibited further reduction in IBV’s copies in comparison with the pure IVM-treated group. Conclusions: PLs loaded with IVM are an innovative and potentially effective way to inhibit IBV. Full article

A large proportion of the lowest annual farm profits in the United Kingdom in recent years comes from lowland and Less Favoured Area (LFA) beef and sheep farms. Benchmarking the performance of a business through routine data collection can provide the information needed to make changes to enterprise management and performance. Key performance indicators (KPIs) are globally recognised measures that can provide farmers with this capability. However, it is largely unknown if there are specific KPIs relating to livestock production that have a significant effect on financial performance. The aim of this study was to determine whether KPIs could be used as predictors of financial performance (gross margin, GM), on suckler beef and sheep farms in England and Wales. This was completed using data from the Farm Business Survey (FBS), which is the largest stratified financial survey of its kind in the UK. Following data extraction, multiple linear regression models were developed for four enterprise types: LFA suckler beef, lowland suckler beef, LFA ewe and lowland ewe. Several KPIs were significantly associated with gross margin per head in all enterprise types. KPIs that were positively associated with GM were measures of livestock productivity, which were lambs per breeding stock and calves per cow. The increased expenditure on concentrate feed had a significantly negative association within all enterprise types, except for LFA suckler beef enterprises, where cow mortality had the greatest significantly negative association. This is the first study to demonstrate the influence livestock production KPIs have on the financial performance of suckler beef and sheep enterprises in both England and Wales, highlighting the importance of routine data collection and benchmarking. Full article

Background/Objectives: The necessity for ground-breaking treatments for Gram-negative infections is evident. The World Health Organization, the Infectious Diseases Society of America, and the European Commission have highlighted the critical insufficiency of efficient antibiotics, urging pharmaceutical businesses to manufacture new antibiotics. Therefore, developing new antibiotics with broad efficacy against Gram-negative pathogens is essential. Thus, this research aimed to evaluate the safety and effectiveness of cefiderocol in treating multidrug-resistant Gram-negative bacterial infections at the International Medical Center (IMC), Jeddah, Saudi Arabia. Methods: A retrospective analysis was conducted on patients treated from January 2021 to February 2023. Thirteen case groups treated with cefiderocol were compared to twenty control groups treated with other antibiotics. Results: The results indicated no statistically significant differences in ICU stay, comorbidity indices, or mortality rates between the two groups. Cefiderocol showed high clinical and microbiological cure rates, despite the severity of the patients’ conditions. Carbapenem-resistant Klebsiella pneumoniae and difficult-to-treat resistance Pseudomonas aeruginosa were the most prevalent pathogens in the case and control group, respectively. Two patients treated with cefiderocol developed Clostridioides difficile infection, emphasizing the need for close monitoring of potential adverse effects. Conclusions: The results of this study support cefiderocol as a viable alternative for managing serious infections instigated by multidrug-resistant Gram-negative bacteria. Full article

Combined PM_2.5 and O₃ pollution in China has caused negative health impacts on residents and reduced crop yields. The quantitative assessment and prediction of these impacts could provide a scientific basis for policy development. This study assessed the nationwide premature mortality, health effects, and crop damage attributable to PM_2.5 and O₃ pollution in 2019, and projected the associated health and agricultural losses under a business-as-usual (BAU) scenario for 2025. The economic benefits of improving air quality under different policy scenarios, including the 14th Five-Year Plan (FFP), Secondary Standard Limit (SSL), and Primary Standard Limit (PSL), were also explored. The results showed PM_2.5 pollution in 2019 resulted in 246,000 all-cause premature deaths and the economic health loss was RMB 196.509 billion. Similarly, O₃ pollution caused 186,300 premature deaths and the economic health loss was RMB 155.807 billion. O₃ pollution has led to a loss of 28.5241 million tonnes of crop production and an economic loss of RMB 62.268 billion. Compared with 2019, the avoidable premature deaths from PM_2.5 under different scenarios in 2025 were 50,600, 43,000, and 200,300 cases, respectively, exceeding the number of avoided premature deaths from O₃ pollution. Compared with the BAU, reducing PM_2.5 under different scenarios could generate economic benefits of RMB 70.178 billion, RMB 60.916 billion, and RMB 229.268 billion. Furthermore, the FFP scenario outperformed the SSL in mitigating winter wheat production losses caused by O₃ pollution. These results provide important scientific support for the development and evaluation of future comprehensive pollution control measures for PM_2.5 and O₃. Full article