Search Results (3,938)

Most existing deep learning models for remote sensing scene classification (RSSC) adopt an offline learning paradigm, where all classes are jointly optimized on fixed-class datasets. In dynamic real-world scenarios with streaming data and emerging classes, such paradigms are inherently prone to catastrophic forgetting when models are incrementally trained on new data. Recently, a growing number of class-incremental learning (CIL) methods have been proposed to tackle these issues, some of which achieve promising performance by rehearsing training data from previous tasks. However, implementing such strategy in real-world scenarios is often challenging, as the requirement to store historical data frequently conflicts with strict memory constraints and data privacy protocols. To address these challenges, we propose a novel memory-efficient feature-replay CIL framework (FR-CIL) for RSSC that retains compact feature embeddings, rather than raw images, as exemplars for previously learned classes. Specifically, a progressive multi-scale feature enhancement (PMFE) module is proposed to alleviate representation ambiguity. It adopts a progressive construction scheme to enable fine-grained and interactive feature enhancement, thereby improving the model’s representation capability for remote sensing scenes. Then, a specialized feature calibration network (FCN) is trained in a transductive learning paradigm with manifold consistency regularization to adapt stored feature descriptors to the updated feature space, thereby effectively compensating for feature space drift and enabling a unified classifier. Following feature calibration, a bias rectification (BR) strategy is employed to mitigate prediction bias by exclusively optimizing the classifier on a balanced exemplar set. As a result, this memory-efficient CIL framework not only addresses data privacy concerns but also mitigates representation drift and classifier bias. Extensive experiments on public datasets demonstrate the effectiveness and robustness of the proposed method. Notably, FR-CIL outperforms the leading state-of-the-art CIL methods in mean accuracy by margins of 3.75%, 3.09%, and 2.82% on the six-task AID, seven-task RSI-CB256, and nine-task NWPU-45 datasets, respectively. At the same time, it reduces memory storage requirements by over 94.7%, highlighting its strong potential for real-world RSSC applications under strict memory constraints. Full article

Well-designed randomized controlled trials (RCTs) have demonstrated safe and effective use of oral antimicrobial therapy for bone and joint infections. Application of data for implementation into real-world practice, however, has inherent challenges. This retrospective cohort study compared real-world use of intravenous versus oral antimicrobial therapy in bone and joint infections within a large healthcare system comprising both academic and community medical centers. The primary outcome was the proportion of treatment failure. Key secondary outcomes included the proportion of patients with logistical failure and risk factors associated with treatment and logistical failure. Among 166 patients included, 136 (82%) and 30 (18%) received predominantly intravenous and oral antimicrobial therapy, respectively. Treatment failure occurred in (77/121) 64% versus (18/25) 72% of patients in the intravenous and oral antimicrobial groups (p = 0.491; OR, 1.38; 95% CI, 0.56–3.33). Logistical failure occurred in 29% versus 47% of patients in the intravenous and oral antimicrobial groups (p = 0.150; OR, 1.93; 95% CI 0.79–4.70). Risk factors for treatment failure included peripheral vascular disease (OR, 2.61; 95% CI 1.02–7.80) and higher Charlson Comorbidity Index scores (OR, 1.18; 95% CI 1.04–1.36). Similar to previously published RCTs, treatment failure appeared comparable between groups; however, oral antimicrobial therapy was overall underutilized. Full article

Purpose: Cachexia, characterized by involuntary weight loss, muscle wasting, and metabolic dysfunction, is prevalent in advanced cancer and chronic illnesses. Despite its impact, outpatient treatment models in the U.S. remain limited and unstandardized. Here, we aim to describe the structure, implementation, patient characteristics, and real-world clinical trajectories of a multidisciplinary clinic for cancer cache as well as other forms of unintentional weight loss clinic within an academic endocrinology practice. Methods: We conducted a retrospective observational cohort study of 103 patients referred to a single-center unintentional weight loss clinic over five years. Patients received comprehensive assessments (weight trajectory, nutrition status, 5× sit-to-stand test, handgrip strength) and personalized interventions including nutrition counseling, resistance training, and pharmacologic therapies. Results: Among 103 patients (median age 69.7 years; 53% male), 64% had cancer, while 36% were referred for non-malignant causes of weight loss or cachexia. Reduced appetite or food intake was reported in 43%, and functional impairment was common, with low handgrip strength in 47% and impaired 5× sit-to-stand performance in 79% of assessed patients. Systemic abnormalities were frequent, including elevated hs-CRP (57%), elevated neutrophil-to-lymphocyte ratio (43%), and hypoalbuminemia (26%). Among patients with available paired follow-up data, the median rate of weight change shifted from −0.5 kg/month prior to enrollment to 0.0 kg/month three months after the initial visit (p < 0.0001). Five-times sit-to-stand performance improved modestly at three months (p = 0.042), while handgrip strength was unchanged. Half of patients that engaged with the clinic returned for at least follow-up, but there was no identifiable difference between the population of patients that returned versus those that did not. Conclusions: A structured, multidisciplinary unintentional weight loss clinic in an endocrinology setting was associated with stabilization of weight and modest changes in physical function in this single-center cohort among patients who engaged in follow-up. These findings highlight the successful implementation of integrated outpatient care models and provide practice-based context for future interventions and therapeutic evaluations. Full article

Background: BH3 mimetics (such as venetoclax and navitoclax) are increasingly investigated in the context of the “one-two punch” anticancer strategy, wherein senescence-inducing therapies are combined with senolytic clearance. However, real-world pharmacovigilance evidence describing hematologic adverse event (AE) patterns and serious outcomes for venetoclax versus navitoclax in such combination settings remains limited. This study aims at providing an expectation based on the current reporting of the safety implications of senolytics combined with senescence-inducing therapy in clinical practice. Methods: We analyzed de-duplicated U.S. FDA Adverse Event Reporting System (FAERS) reports retrieved on 1 August 2025. Venetoclax reports (Q2 2016–Q2 2025) were categorized as monotherapy or combination with senescence-inducing chemotherapy (predefined based on published evidence of therapy-induced senescence [TIS]). Hematologic AEs were grouped into three categories (isolated low WBC, isolated low platelet count, and multi-lineage cytopenia). Disproportionality analyses were conducted using the Reporting Odds Ratio (ROR) and Proportional Reporting Ratio (PRR) with 95% CIs and chi-squared testing. Navitoclax reports were analyzed descriptively due to limited volume. Results: A total of 47,508 venetoclax reports were included (34,485 monotherapy; 13,023 combination). Compared with monotherapy, combination therapy showed disproportionate reporting signals (ROR/PRR; reflecting reporting disproportionality rather than incidence or causal risk) for low WBC (ROR 2.87, PRR 2.59) and multi-lineage cytopenias (ROR 3.54, PRR 2.94), while isolated low platelet count was under-represented (ROR 0.31, PRR 0.32). For outcomes, combination therapy demonstrated higher reporting signals for life-threatening outcomes (ROR 7.06, PRR 6.56), hospitalization (ROR 1.74, PRR 1.39), and other outcomes (ROR 2.36, PRR 1.57), while death (ROR 0.55, PRR 0.65) and non-serious outcomes (ROR 0.26, PRR 0.29) were proportionally less reported (all p < 0.001). Navitoclax had 172 reports; hematologic cytopenias and serious outcomes were frequent, but analyses were descriptive only. Conclusions: In FAERS, venetoclax combined with senescence-inducing chemotherapy shows stronger reporting signals for leukopenia and multi-lineage cytopenias and for several serious outcome categories compared with monotherapy. These reporting patterns highlight the need for further care in terms of clinical implementation of the currently investigated senolytics prior to the consideration of the “one-two punch” strategy. Full article

BCI biosensors enable continuous monitoring of neural activity, but existing systems face challenges in scalability, latency, and reliable integration with cloud infrastructure. This work presents a cloud-aware, real-time cognitive grid architecture for multimodal BCI biosensors, validated at the system level through a full physical prototype. The system integrates the BioAmp EXG Pill for signal acquisition with an RP2040 microcontroller for local preprocessing using edge-resident TinyML deployment for on-device feature/inference feasibility coupled with environmental context sensors to augment signal context for downstream analytics talking to the external world via Wi-Fi/4G connectivity. A tiered data pipeline was implemented: SD card buffering for raw signals, Redis for near-real-time streaming, PostgreSQL for structured analytics, and AWS S3 with Glacier for long-term archival. End-to-end validation demonstrated consistent edge-level inference with bounded latency, while cloud-assisted telemetry and analytics exhibited variable transmission and processing delays consistent with cellular connectivity and serverless execution characteristics; packet loss remained below 5%. Visualization was achieved through Python 3.10 using Matplotlib GUI, Grafana 10.2.3 dashboards, and on-device LCD displays. Hybrid deployment strategies—local development, simulated cloud testing, and limited cloud usage for benchmark capture—enabled cost-efficient validation while preserving architectural fidelity and latency observability. The results establish a scalable, modular, and energy-efficient biosensor framework, providing a foundation for advanced analytics and translational BCI applications to be explored in subsequent work, with explicit consideration of both edge-resident TinyML inference and cloud-based machine learning workflows. Full article

The rapid incorporation of solar energy (PV) systems into electrical grids has increased the demand for accurate short-term forecasts to ensure stability and improve processes. Although hybrid artificial intelligence (AI) models are increasingly being suggested to address this challenge, there is a lack of systematic compilation of their structures, effectiveness, and readiness for use in real-world applications. This paper provides a detailed analysis of 58 peer-reviewed articles (2020–2025) focused on hybrid models for short-term (1–24 h) solar photovoltaic power forecasting. We propose an innovative classification that groups hybrids into four categories: AI-AI (28%), AI with optimization (21%), decomposition-based (17%), and image-based (7%). Our research indicates that weather conditions (34%) and historical photovoltaic energy records (32%) are the most frequent inputs, and that optimized hybrids and those using decomposition achieve the best balance between effectiveness and computational efficiency. From a geographical perspective, the study focuses mainly on the United States (29%) and China (22%), suggesting that more extensive climate validation is crucial. Essentially, we have identified ongoing obstacles to implementation, such as high computational costs, data quality issues, and gaps in interpretation. In addition, we present a plan for future research focusing on hybrid architectures that are lightweight, understandable, and interactive with the grid. This analysis provides a thorough assessment of the current landscape and a strategic framework to guide the creation of operational forecasting systems capable of supporting highly solar-integrated grids. Full article

This study explores classical and Bayesian estimation for the two-parameter Chen distribution with randomly censored data, where censoring times follow an independent two-parameter Chen distribution with separate shape and scale parameters. We first derive the maximum likelihood estimators of the unknown parameters, together with their asymptotic variances and credible intervals, and further adopt the method of moments, L-moments and least squares methods for classical estimation. Under the generalized entropy loss function and inverse gamma priors, Bayesian estimation is implemented via Gibbs sampling, with the highest posterior density credible intervals of parameters constructed accordingly. We also investigate the estimation of key reliability and lifetime characteristics of the distribution, and conduct Monte Carlo simulations to compare the performance of all aforementioned estimation methods. Finally, two real-world CMAPSS jet engine lifetime datasets from NASA are applied to validate the practical effectiveness of the proposed estimation approaches, demonstrating the enhanced flexibility of the Chen distribution compared to the exponential distribution in fitting aerospace-related censored data, given the marginal p-values in the K-S tests. Full article

In Germany, a shortage of public charging stations for the significantly increasing number of electric cars exists. This shortage, along with the associated range anxiety, poses a hurdle for the market entry of electric cars. In addition, the construction of public charging stations for electric vehicles is outside the control of vehicle owners. Accelerating this process is challenging due to regulatory and other considerations. This paper presents a novel scientific and technical approach to complement existing public charging infrastructure by integrating privately owned charging stations into a publicly accessible ecosystem. The core of our work explores the integration of private charging stations into the application ecosystem, clarifying the reservation process and addressing potential challenges. Furthermore, we provide a comprehensive overview of the features related to routing and locating near charging stations. We examine the potential challenges that may arise during the practical implementation of the proposed system. The technical feasibility of the approach is validated through implementation and simulation, demonstrating the practical applicability of the proposed system and its ability to support real-world usage scenarios. The results indicate that such a system has significant potential to enhance the accessibility and usability of charging infrastructure, thereby promoting sustainable mobility and lowering the entry barriers for electric vehicles. By providing integrated charging information, reservation functionality, and route planning, the proposed solution contributes to increasing user acceptance and supporting the transition toward more environmentally friendly transportation. Full article

As the manufacturing industry continues to advance its digital transformation, intelligent sensing technology has become a key support for achieving precise, efficient and automated quality control. However, current production line monitoring systems predominantly rely on fixed and costly monitoring equipment and sensors, lacking flexible and interactive first-person perspective perception approaches centered on on-site operators. Meanwhile, factory process monitoring often depends solely on visual expression rather than balancing the capabilities of the simulation model and visual state detection, leading to delayed responses to abnormal systems and hindering the adjustment strategy feedback. To address these limitations, this study provides wearable sensing for key workers, enriching the state perception capabilities in industrial scenarios. Furthermore, to achieve dynamic model and real-time visual representation of production line operations, a multi-source information-enhanced Petri nets model is proposed in terms of engineering and user-friendliness. With the solid mathematical basics of the Petri nets and the enriched human–machine data from the product line, this method provides an intuitive, dynamic and accurate reflection of the production system’s real-time operational status, offering a scientific and reliable basis for operational decision-making. The proposed approach has been implemented in a real-world production system for reinforced concrete civil defense doors, and this engineering application can also be extended to many other scenarios. Full article

Checkout service is crucial in ensuring customer satisfaction and enhancing retail efficiency. In recent years, self-checkout has become increasingly popular in modern retail operations. However, despite its growing adoption, there is limited quantitative evidence on its effectiveness in reducing operational costs and improving overall efficiency. In this study, a discrete-event simulation model based on real-world scenarios of a retail store in Taoyuan City, Taiwan, was developed using ARENA (version 16) simulation software. Four checkout scenarios were modeled and compared through statistical tests to evaluate checkout performance. The results showed that the proposed self-checkout model with improved service time enhanced operational efficiency and contributed to reducing operational costs. These findings suggest that retail managers should implement strategic measures to optimize self-checkout operations to achieve efficient and cost-effective store performance. Finally, practical and managerial implications are discussed at the end of the study. Full article

Cardiovascular imaging is a cornerstone of modern cardiology, yet its clinical impact is limited by operator dependence, inter-observer variability, time-consuming workflows, and unequal access to advanced expertise. Artificial intelligence (AI), particularly machine learning and deep learning, offers new opportunities to overcome these limitations. This review aims to summarize current and emerging AI applications in cardiovascular imaging and to evaluate their potential clinical value in precision diagnostics and decision support. This narrative review synthesizes clinically relevant literature on AI applications across major cardiovascular imaging modalities, including echocardiography, cardiovascular magnetic resonance, cardiac computed tomography, and nuclear cardiology. Evidence was analyzed with a focus on AI-enabled acquisition support, image segmentation, quantitative and functional assessment, workflow automation, and risk stratification, alongside key methodological and implementation considerations. Across imaging modalities, AI-driven approaches have demonstrated improved reproducibility, efficiency, and scalability of cardiovascular imaging workflows. Automated algorithms reduce operator dependence, facilitate standardized extraction of imaging biomarkers, and support advanced functional assessment and prognostic stratification. Recent developments in video-based, temporal, and multimodal models further expand AI capabilities from technical automation toward integrated disease phenotyping and personalized clinical decision support. However, translation into routine practice remains limited by heterogeneous datasets, insufficient external validation, algorithmic bias, limited interpretability, and challenges related to regulatory approval and workflow integration. Artificial intelligence has the potential to reshape cardiovascular imaging into a more efficient, reproducible, and patient-centered precision medicine tool. Real-world clinical impact will depend on outcome-driven evaluation, robust external validation, multimodal data integration, and human-in-the-loop implementation strategies that ensure safe, equitable, and clinically meaningful adoption. Full article

Out-of-hospital cardiac arrest (OHCA) remains a leading cause of sudden death worldwide, with wide variation in reported incidence and outcomes driven by heterogeneity in registries, emergency medical services (EMS) organization, and case definitions. Despite substantial advances in resuscitation systems, survival after EMS-treated OHCA generally remains below 10%, and outcomes are critically time dependent. Delays in emergency call activation, bystander cardiopulmonary resuscitation (CPR), and—most importantly—early defibrillation are associated with a rapid decline in return of spontaneous circulation and favorable neurological recovery. This narrative review synthesizes current evidence and implementation strategies aimed at reducing “time-to-CPR” and “time-to-shock,” with a specific focus on public-access defibrillation (PAD) as a tool to mitigate avoidable delay. Randomized trials and large registry studies consistently demonstrate that automated external defibrillator (AED) use before EMS arrival is a key determinant of survival in patients with shockable rhythms. However, the real-world effectiveness of PAD remains limited by suboptimal AED placement, restricted 24/7 accessibility, low public awareness, and underutilization driven by fear and lack of confidence. We compare different PAD delivery models—including EMS-based, police and first-responder-based, and fully integrated community systems—and summarize evidence supporting targeted, high-yield AED deployment and cost-effectiveness. In addition, we review emerging strategies to reduce avoidable delay and strengthen the early links of the chain of survival, such as school-based training programs, smartphone- and SMS-based citizen-responder networks, improved dispatch recognition of cardiac arrest (including artificial intelligence–supported tools), and drone-enabled AED delivery. Across these approaches, patient benefit critically depends on system integration, alert performance, and true AED accessibility. Finally, we describe the Italian “Progetto Vita” experience as a community-integrated model explicitly designed to minimize avoidable delay through widespread AED deployment, lay responder training, and real-time integration with EMS. We conclude by outlining future priorities, including the development of robust national OHCA registries and scalable solutions for the high burden of cardiac arrests occurring at home, such as population-level deployment of low-cost, ultra-portable AEDs. Full article

The well-known Heath–Jarrow–Morton (HJM) framework provides a universal and efficacious instrument for modeling the stochastic evolution of an entire yield curve by explaining the interest rate dynamics in continuous time under no-arbitrage conditions. Existing implementations involve exponentially weighted function spaces as theoretical settings for the former stochastic evolution. While the choice of weight can have a drastic effect on model calibration and subsequent forecasting, it cannot be estimated from market data and does not allow for any objective interpretation. The proposed approach does not have this shortcoming as it adopts a suitably designed unweighted function space. The HJM equation is discretized using a finite difference approach. The resulting semiparametric model is then calibrated on real-world yield data with a new type of functional principal component analysis (PCA)-based approach. Backtesting and benchmarking are conducted against the one-factor Vasicek model using historical data to illustrate its simulation capabilities for prediction and uncertainty quantification. Additionally, in contrast to widely studied US treasuries, negative interest rates are observed for AAA Euro Bonds during the sample period employed for this study. Accordingly, the framework allows for the possibility of negative yields. Full article

Background: Contemporary guidelines recommend rapid initiation of four classes of guideline-directed medical therapy (GDMT) for heart failure (HF) with reduced ejection fraction (HFrEF); however, real-world persistence, adherence, and dose optimization remain suboptimal. Methods: We analysed a predefined subregistry within the prospective Cardiology Research Dubrava (CaRD) registry, a real-world HF registry at a tertiary centre that includes patients across the ejection-fraction spectrum in whom contemporary HF therapy, including sodium-glucose cotransporter 2 inhibitors (SGLT2i), is introduced or optimised in routine practice. For this analysis, we included patients with newly diagnosed HFrEF (left ventricular ejection fraction (LVEF) ≤ 40%) who were discharged on all four GDMT classes; 167 of 179 patients with newly diagnosed HFrEF during the study period had an available 6-month medication assessment and comprised the final analytic cohort. The four GDMT pillars (beta-blocker; angiotensin-converting enzyme inhibitor (ACEi), angiotensin receptor blocker (ARB), or angiotensin receptor-neprilysin inhibitor (ARNI); mineralocorticoid receptor antagonist (MRA); and SGLT2i) were initiated within 4 days when clinically feasible. Medication adherence and target-dose attainment were assessed at 6, 12, and 24 months using a structured self-report questionnaire. Major adverse events (MAE) and all-cause mortality were recorded over 24 months. Patients were classified as adherent if they reported regular intake (≥80% of prescribed doses) of all four drug classes at 6 months; otherwise, they were classified as nonadherent. Results: Among the 167 analysed patients (median age 64 years, 74% men, median LVEF 30%), regular adherence at 6, 12, and 24 months was 65%, 55%, and 59% for beta-blockers; 66%, 50%, and 49% for ACEi/ARB/ARNI; 62%, 52%, and 49% for MRAs; and 84%, 57%, and 68% for SGLT2i. Target doses were achieved in 25–33% for beta-blockers, 42–50% for ACEi/ARB/ARNI, and 73–78% for MRAs. At 24 months, 56 survivors (37%) were adherent to all four drug classes. Over 24 months, all-cause mortality was 9.0% and MAE 18.6%, occurring less frequently in adherent vs. nonadherent patients (mortality 0% vs. 13.5%; MAE 8.9% vs. 23.4%). Conclusions: In this real-world, non-randomized HFrEF subregistry, in-hospital initiation of quadruple GDMT was feasible, yet maintaining long-term adherence and achieving target doses remained challenging. These data underscore the gap between guideline recommendations and routine practice and support structured follow-up and protocol-driven titration to optimize implementation. Full article

Background: Post-traumatic stress disorder (PTSD) affects millions globally, with 40–50% of patients not responding adequately to first-line treatments. Prism neurofeedback, an FDA-cleared electroencephalography (EEG)-based system targeting amygdala-derived biomarkers, has demonstrated efficacy in randomized controlled trials (RCTs) and multicenter studies. Real-world implementation data from community clinical practice remain limited. Objective: To evaluate clinical outcomes and patient-developed self-regulation strategies of Prism neurofeedback in patients with PTSD in community clinical practice. Methods: Retrospective case series of 28 consecutive patients with PTSD treated with Prism neurofeedback in a community psychiatry practice. The primary outcome was change in PTSD Checklist for DSM-5 (PCL-5) from baseline to end of treatment. Results: Twenty-one of 28 patients (75.0%) completed treatment. Mean PCL-5 reduction was 37.0 ± 18.2 points (Cohen’s d = 2.03). Response rates were 100% for any improvement and 90.5% for clinically significant improvement (≥10-point reduction). Five patients (23.8%) achieved excellent response with ≥50-point reduction. Limited follow-up data (1–3 months post-treatment) were available for three patients; two of three (67%) exceeded their end-of-treatment gains. Four patients receiving booster sessions showed continued improvement. Limitations: The uncontrolled, retrospective design precludes causal attribution of improvements to the intervention versus placebo effects or regression to the mean. The 25% early discontinuation rate may introduce attrition bias. Durability data are available for only three patients. Conclusions: This case series provides real-world evidence supporting the feasibility and potential clinical utility of Prism neurofeedback in community practice, with outcomes comparable to controlled studies and preliminary evidence of durable treatment effects. These findings complement existing RCT evidence by demonstrating successful implementation outside research settings. Full article