Search Results (424)

Deep generative models trained on sensitive data pose significant privacy risks, yet enforcing differential privacy (DP) in high-dimensional generators often leads to severe utility degradation. We propose Differentially Private Vector-Quantized Generation (DP-VQG), a three-stage generative framework that introduces a discrete latent bottleneck as the interface for privacy preservation. DP-VQG separates geometric structure learning, differentially private discrete latent projection, and non-private prior modeling, ensuring that privacy-induced randomness operates on a finite codebook aligned with the decoder’s effective support. This design avoids off-support degradation while providing formal end-to-end DP guarantees through composition and post-processing. We provide a theoretical analysis of privacy and utility, including explicit bounds on privacy-induced distortion. Empirically, under the privacy budget of $\epsilon =10$, DP-VQG attains Fréchet Inception Distance (FID) scores of 18.21 on MNIST and 77.09 on Fashion-MNIST, surpassing state-of-the-art differentially private generative models of comparable scale. Moreover, DP-VQG produces visually coherent samples on high-resolution datasets such as Flowers102, Food101, CelebA-HQ, and Cars, demonstrating scalability beyond prior end-to-end DP generative approaches. Full article

Background/Objectives: (P)rehabilitation, comprising structured exercise, nutritional optimisation, and/or psychological support delivered pre- or postoperatively, has demonstrated efficacy in improving outcomes across the cancer care continuum. However, access remains limited. Technology-enabled (p)rehabilitation offers a novel solution with the potential to enhance equity and continuity of care. This systematic review aimed to evaluate the efficacy of technology-enabled (p)rehabilitation on perioperative and patient-reported outcomes among individuals undergoing thoracic and/or abdominopelvic cancer surgery. Methods: Six databases were search from inception to October 2024. Eligible studies were randomised controlled trials (RCTs) comparing technology-enabled (p)rehabilitation with usual care, placebo, or non-technology-based interventions in adults undergoing thoracic and/or abdominopelvic cancer surgery. Outcomes included postoperative complications, hospital readmissions, hospital length of stay (LOS), quality of life (QoL), pain, anxiety, depression, fatigue, distress, and satisfaction. Higher scores indicated improved QoL or worse symptom severity. Risk of bias was assessed using the revised Cochrane tool, and evidence strength was determined using GRADE methodology. Relative risks (RR) and mean differences (MD) were calculated using random-effects meta-analysis. Results: Seventeen RCTs (18 publications, n = 1690) were included. Trials most commonly evaluated application-based platforms (n = 8) and the majority exhibited some risk of bias. Technology-enabled (p)rehabilitation was associated with a significant reduction in LOS (MD = 1.33 days; 95% CI: 0.59–2.07; seven trials), and improvements in pain (MD = 6.12; 95% CI: 3.40–8.84; four trials), depression (MD = 2.82; 95% CI: 0.65–4.99; five trials), fatigue (MD = 10.10; 95% CI: 6.97–13.23; three trials) and distress (MD = 1.23; 95% CI: 0.30–2.16; single trial) compared with controls. Conclusions: Technology-enabled (p)rehabilitation shows promise in reducing LOS and improving selected patient-reported outcomes following thoracic and abdominopelvic cancer surgery. Although evidence is limited due to the small number of studies, modest sample sizes, methodological heterogeneity, and intervention variability, the overall findings justify further investigation. Large-scale, adequately powered clinical trials are required to confirm efficacy and guide clinical effectiveness and implementation studies. Full article

Traditional Turkish marbling (Ebru) art is an intangible cultural heritage characterized by highly asymmetric, fluid, and non-reproducible patterns, making its long-term preservation and large-scale dissemination challenging. It is highly sensitive to environmental conditions, making it enormously difficult to mass produce while maintaining its original aesthetic qualities. A data-driven generative model is therefore required to create unlimited, high-fidelity digital surrogates that safeguard this UNESCO heritage against physical loss and enable large-scale cultural applications. This study introduces a deep generative modeling framework for the digital reconstruction of traditional Turkish marbling (Ebru) art using a Deep Convolutional Generative Adversarial Network (DCGAN). A dataset of 20,400 image patches, systematically derived from 17 original marbling works, was used to train the proposed model. The framework aims to mathematically capture the asymmetric, fluid, and stochastic nature of Ebru patterns, enabling the reproduction of their aesthetic structure in a digital medium. The generated images were evaluated using multiple quantitative and perceptual metrics, including Fréchet Inception Distance (FID), Kernel Inception Distance (KID), Learned Perceptual Image Patch Similarity (LPIPS), and PRDC-based indicators (Precision, Recall, Density, Coverage). For experimental validation, the proposed DCGAN framework is additionally compared against a Vanilla GAN baseline trained under identical conditions, highlighting the advantages of convolutional architectures for modeling marbling textures. The results show that the DCGAN model achieved a high level of realism and diversity without mode collapse or overfitting, producing images that were perceptually close to authentic marbling works. In addition to the quantitative evaluation, expert qualitative assessment by a traditional Ebru artist confirmed that the model reproduced the organic textures, color dynamics, and compositional asymmetrical characteristic of real marbling art. The proposed approach demonstrates the potential of deep generative models for the digital preservation, dissemination, and reinterpretation of intangible cultural heritage recognized by UNESCO. Full article

Background/Objectives: Postoperative pulmonary complications (PPCs) remain frequent and increase morbidity, mortality, and resource use. Preoperative risk scores (ARISCAT, NSQIP-derived calculators) use mostly static variables and may miss the dynamic perioperative host response preceding respiratory deterioration or infection. We address the gap in clinically interpretable syntheses of perioperative blood biomarker trajectories that distinguish infectious from non-infectious PPCs and clarify bedside-ready versus exploratory markers. Methods: We conducted a narrative review with a structured Medline search (inception to 1 November 2025) plus reference screening. We included English-language adult surgical studies (observational or interventional) evaluating perioperative blood biomarkers in relation to PPCs or postoperative pulmonary infection; case reports, editorials, and reviews were excluded. No formal risk-of-bias assessment or quantitative meta-analysis was performed. Results: Across 298 cited publications, serial patterns of routinely available biomarkers (C-reactive protein, procalcitonin, lactate, albumin, and leukocyte-derived indices) were most consistently associated with PPC risk and helped separate expected postoperative inflammation from evolving infection when interpreted longitudinally rather than as single values. Mechanistic biomarkers (cytokines/immune-function assays, endothelial injury and coagulation/fibrinolysis markers, oxidative stress indicators) add biological insight but are limited by assay availability, heterogeneous sampling windows, and absent standardized cut-offs. Omics signatures and machine learning models combining biomarker kinetics with clinical variables are promising but require prospective, transportable validation. Conclusions: Key barriers to implementation include biological variability, non-specificity across postoperative syndromes, heterogeneous sampling windows, and lack of standardized cut-offs. Integrating multimarker panels into validated, dynamic predictive frameworks represents a promising direction for perioperative precision medicine. Full article

A universal approach to managing a person’s digital identity may be the single most important advancement to the Internet since its inception, promising the seamless flow of information, averting cybercrime, eliminating login credentials, and restoring privacy and trust through greater control of one’s identity online. However, this advancement brings significant risks, especially regarding personal privacy. It demands the meticulous development of digital identity infrastructure that balances robust data security measures with ethical handling of sensitive information, thereby safeguarding against misuse and unauthorised access. Currently, a consolidated vision for digital identity implementation remains unresolved, and aligning the different stakeholders’ motives and expectations is a challenging task. This article reviews and analyses the perspectives and expectations of four key stakeholder groups—government, business, academia, and consumers—regarding a digital identity ecosystem, aiming to increase trust in an eventual design framework. Using an online survey stratified across government, business, academia, and consumers, we identify areas of alignment and divergence regarding privacy, trust, usability, and governance expectations. We then encode these stakeholder expectations into a layered conceptual structure and illustrate its use as metadata for context-layered retrieval-augmented generation (RAG) in digital identity scenarios. Full article

The structural analysis of paper fibers is vital for the noninvasive classification and conservation of traditional handmade paper in cultural heritage. Although digital still cameras (DSCs) offer a low-cost and noninvasive imaging solution, their inferior image quality compared to white-light confocal microscopy (WCM) limits their effectiveness in fiber classification. To address this modality gap, we propose an unpaired image-to-image translation approach using cycle-consistent adversarial networks (CycleGANs). Our study targets a multifiber setting involving kozo, mitsumata, and gampi, using publicly available domain-specific datasets. Generated WCM-style images were quantitatively evaluated using peak signal-to-noise ratio, structural similarity index measure, mean absolute error, and Fréchet inception distance, achieving 8.24 dB, 0.28, 172.50, and 197.39, respectively. Classification performance was tested using EfficientNet-B0 and Inception-ResNet-v2, with F1-scores reaching 94.66% and 98.61%, respectively, approaching the performance of real WCM images (99.50% and 98.86%) and surpassing previous results obtained directly from DSC inputs (80.76% and 84.19%). Furthermore, Grad-CAM visualization confirmed that the translated images retained class-discriminative features aligned with those of the actual WCM inputs. Thus, the proposed CycleGAN-based image conversion effectively bridges the modality gap, enabling DSC images to approximate WCM characteristics and support high-accuracy paper fiber classification, which is a practical alternative for noninvasive material analysis. Full article

Background and Objectives: The integrated stress response (ISR) is a convergent node in neurodegeneration. We systematically mapped open-access mammalian in vivo evidence for synthetic ISR modulators, comparing efficacy signals, biomarker engagement, and safety across mechanisms and disease classes. Methods: Following PRISMA 2020, we searched PubMed (MEDLINE), Embase, and Scopus from inception to 22 September 2025. Inclusion required mammalian neurodegeneration models; synthetic ISR modulators (eIF2B activators, PERK inhibitors or activators, GADD34–PP1 ISR prolongers); prespecified outcomes; and full open access. Extracted data included model, dose and route, outcomes, translational biomarkers (ATF4, phosphorylated eIF2α), and safety. Results: Twelve studies met the criteria across tauopathies and Alzheimer’s disease (n = 5), prion disease (n = 1), amyotrophic lateral sclerosis and Huntington’s disease (n = 3), hereditary neuropathies (n = 2), demyelination (n = 1), and aging (n = 1). Among interpretable in vivo entries, 10 of 11 reported benefit in at least one domain. By class, eIF2B activation with ISRIB was positive in three of four studies, with one null Alzheimer’s hAPP-J20 study; PERK inhibition was positive in all three studies; ISR prolongation with Sephin1 or IFB-088 was positive in both studies; and PERK activation was positive in both studies. Typical regimens included ISRIB 0.1–2.5 mg per kg given intraperitoneally (often two to three doses) with reduced ATF4 and phosphorylated eIF2α; oral GSK2606414 50 mg per kg twice daily for six to seven weeks, achieving brain-level exposures; continuous MK-28 delivery at approximately 1 mg per kg; and oral IFB-088 or Sephin1 given over several weeks. Safety was mechanism-linked: systemic PERK inhibition produced pancreatic and other exocrine toxicities at higher exposures, whereas ISRIB and ISR-prolonging agents were generally well-tolerated in the included reports. Conclusions: Directional ISR control yields consistent, context-dependent improvements in behavior, structure, or survival, with biomarker evidence of target engagement. Mechanism matching (down-tuning versus prolonging the ISR) and exposure-driven safety management are central for translation. Full article

Maintaining the structural integrity of conveyor belts is essential for safe and reliable mining operations. However, these belts are susceptible to longitudinal tearing and surface degradation from material impact, fatigue, and deformation. Many computer vision-based inspection methods are inefficient and unreliable in harsh mining environments characterized by dust and variable lighting. This study introduces a smartphone-driven defect detection system for the cost-effective, geometric inspection of conveyor belt surfaces. Using Apple’s iPhone 12 Pro Max (Apple Inc., Cupertino, CA, USA), the system captures 3D point cloud data from a moving belt with induced damage via the integrated TrueDepth camera. A key innovation is a 3D-to-2D projection pipeline that converts point cloud data into structured representations compatible with standard 2D Convolutional Neural Networks (CNNs). We then propose a hybrid deep learning and machine learning model, where features extracted by pre-trained CNNs (VGG16, ResNet50, InceptionV3, Xception) are classified by ensemble methods (Random Forest, XGBoost, LightGBM). The proposed system achieves high detection accuracy exceeding 0.97 F1 score in the case of all proposed model implementations with TrueDepth F1 score over 0.05 higher than RGB approach. Applied cost-effective smartphone-based sensing platform proved to support near-real-time maintenance decisions. Laboratory results demonstrate the method’s reliability, with measurement errors for defect dimensions within 3 mm. This approach shows significant potential to improve conveyor belt management, reduce maintenance costs, and enhance operational safety. Full article

Advanced prostate cancer, particularly castration-resistant disease, remains challenging to treat due to intratumoral heterogeneity, immune exclusion, and a suppressive tumor microenvironment. Within this ecosystem, cancer-associated fibroblasts shape tumor–stroma communication, but their marked heterogeneity and plasticity complicate classification and make indiscriminate fibroblast depletion potentially ineffective or even harmful. This review summarizes recent progress in fibroblast origins, functional subtypes, and fibroblast-driven mechanisms that promote tumor progression and therapy resistance, as well as emerging therapeutic opportunities in prostate cancer. We conducted a structured literature search of PubMed, ScienceDirect, and major publisher platforms (including Nature and SpringerLink) from database inception to 15 February 2025, supplemented by targeted manual screening of reference lists. Evidence from single-cell/spatial-omics and mechanistic studies indicates that prostate tumors contain multiple fibroblast programs that occupy distinct niches yet can interconvert. Across these studies, it was found that these fibroblasts contribute to immune suppression, extracellular matrix remodeling and stromal barrier formation, angiogenesis, and metabolic support, collectively limiting drug penetration and reinforcing immune evasion; therapeutic pressure can further rewire fibroblast states and resistance-associated signaling. Overall, the literature supports a shift toward function- and subtype-directed intervention rather than “one-size-fits-all” targeting, with promising directions including precision targeting and reversible reprogramming, rational combination strategies, and localized delivery approaches that reduce stromal barriers while preserving tissue homeostasis in high-risk and treatment-refractory prostate cancer. Full article

Sodium–glucose cotransporter 2 (SGLT2) inhibitors have substantially reshaped the management of type 2 diabetes mellitus (T2DM), owing not only to their glucose-lowering properties but also to their consistent cardiovascular and renal protective effects. Beyond their initial metabolic indication, these agents have emerged as disease-modifying therapies across a broad spectrum of cardiometabolic and renal conditions. Building on the clinical success of first-generation SGLT2 inhibitors, such as empagliflozin and dapagliflozin, next-generation SGLT2-based therapies have been developed with the aim of refining pharmacological selectivity, optimizing pharmacokinetic profiles, and expanding therapeutic applicability beyond diabetes. These innovations include dual SGLT1/SGLT2 inhibition, alternative dosing strategies, and molecular designs tailored to specific clinical phenotypes, such as heart failure with preserved ejection fraction (HFpEF) and chronic kidney disease (CKD). This narrative review critically evaluates the evolving landscape of next-generation SGLT2 inhibitors, with a focus on structural and pharmacokinetic innovations, transporter selectivity, glucose-independent mechanisms, and emerging clinical implications. A comprehensive literature search was conducted using PubMed/MEDLINE, Scopus, and Web of Science, encompassing publications from inception to March 2025. Eligible sources included randomized clinical trials, observational studies, meta-analyses, and authoritative reviews published in English. Available evidence indicates that, while conventional SGLT2 inhibitors confer robust and reproducible cardiorenal benefits, newer agents may further extend therapeutic potential through incretin-related effects, modulation of extra-renal pathways, and disease-specific cardiac and renal mechanisms. Nevertheless, evidence supporting incremental clinical benefit beyond established SGLT2 inhibitors remains limited and heterogeneous, particularly for recently developed compounds. Overall safety profiles appear broadly consistent within the class, although long-term data for next-generation agents are still evolving. Key limitations of the current evidence base include reliance on emerging or indirect mechanistic data, heterogeneity in study populations and clinical endpoints, and the relative scarcity of large, outcome-driven trials for newer SGLT2-based therapies. Future research should prioritize mechanism-driven clinical trials, precision-oriented patient stratification, and head-to-head comparative studies to more clearly define the role of next-generation SGLT2 inhibitors in cardiovascular, renal, and metabolic disease management. Full article

Background/Objectives: Type 1 diabetes (T1D) is a common childhood condition with rising global incidence. Because early-onset T1D coincides with key periods of brain maturation, affected children may face neurocognitive risks. This review summarizes current evidence on the neurocognitive impact of pediatric T1D and related clinical implications. Methods: A structured search of PubMed, Scopus, and Web of Science (inception–October 2025) used combinations of terms related to T1D, cognitive outcomes, and brain imaging. Studies involving participants under 18 years that reported cognitive or neuroimaging findings were included. Results: Diabetic ketoacidosis (DKA) at diagnosis is consistently linked with acute and longer-term neurological injury, including reduced brain volume and potential persistent deficits in memory and executive functioning. Severe or recurrent hypoglycemia disproportionately affects the hippocampus, contributing to lasting learning and memory impairments. Chronic hyperglycemia is a major driver of progressive neurocognitive decline; higher HbA1c is associated with smaller brain volumes and poorer executive function, attention, and processing speed. Early-onset disease and longer duration further increase vulnerability. These neurocognitive effects translate into modest reductions in academic performance and quality of life, especially with poor glycemic control. Emerging evidence suggests that continuous glucose monitoring, insulin pumps, and hybrid closed-loop systems improve metabolic stability and may support healthier brain development. Conclusions: T1D children experience subtle but meaningful neurocognitive risks shaped by glycemic extremes and early disease onset. Routine neuropsychological monitoring, strengthened academic support, and wider use of advanced diabetes technologies may help preserve cognitive development. Larger, longitudinal neuroimaging studies are needed to guide targeted neuroprotective strategies. Full article

Background/Objectives: Patients with advanced non-malignant chronic conditions experience illness burdens and palliative care needs comparable to those of oncology patients, yet palliative care is often introduced late. Identifying individuals with potential palliative care needs is complex, and although multiple tools exist, the most appropriate approach for assessing needs in this population remains unclear. This umbrella review aimed to identify and evaluate tools used to systematically assess palliative care in adults with advanced non-malignant chronic conditions, with a specific focus on their content, structure, and psychometric properties. Methods: An umbrella review of systematic reviews was conducted in accordance with the Joanna Briggs Institute (JBI) guidance. Four electronic databases (Cochrane Library, CINAHL, PubMed, and PsycINFO) were searched from inception to 30 June 2025. Eligible systematic reviews were screened, critically appraised, and synthesized narratively. Results: Seven systematic reviews met the inclusion criteria, collectively identifying 35 unique needs-assessment tools. Five tools (SPICT, GSF-PIG, QUICK GUIDE, NECPAL, and P-CaRes) incorporated both general and disease-specific palliative care indicators. At the same time, four (PC-NAT, SPEED, NAT, and IPOS) addressed needs across physical, psychological, social, and spiritual domains. Psychometric data were available for six tools across three reviews. The original NAT and SPICT demonstrated good reliability; however, the Dutch version of the NAT showed poor validity. SPEED and one unnamed palliative care tool showed good reliability, whereas the Surprise Question demonstrated unclear validity. Italian-SPICT and Israeli-NECPAL exhibited strong content validity. Conclusions: Despite limited evidence, the NAT: PD-HF shows particular promise for identifying palliative care needs in patients with heart failure. Tools such as SPICT and NECPAL are widely used and adapted for advanced non-malignant chronic conditions, but further psychometric evaluation is required. Additional studies are needed to clarify the clinical utility of these tools for broader implementation in assessing palliative care needs. Full article

Mountaintop towers are highly exposed to self-initiated upward lightning flashes. Accurate estimation of their effective height—the equivalent flat-ground height yielding the same lightning exposure—is essential for reliable exposure assessment, for interpreting and calibrating measurement data at instrumented mountaintop towers, and for comparison with established protection guidelines. This study applies a two-step numerical framework that couples finite-element electrostatic simulations with a leader-inception and propagation model for representative tower–terrain configurations reflecting reference instrumented mountaintop sites in lightning research. For each configuration, the stabilization field, the minimum background electric field enabling continuous upward leader propagation to the cloud base, is determined, from which effective heights are obtained. The simulated results agree with the analytical formulation of Zhou et al. (within ~10%), while simplified or empirical approaches by Shindo, Eriksson, and Pierce exhibit larger deviations, especially for broader mountains. A normalized analysis demonstrates that the tower-to-mountain slenderness ratio (h/a) governs the scaling of effective height, following a power-law dependence with exponent −0.17 (R² = 0.94). This compact relation enables direct estimation of effective height from geometric parameters alone, complementing detailed leader-inception modeling. The findings validate the proposed physics-based framework, quantify the geometric dependence of effective height for mountaintop towers, and provide a foundation for improving lightning-exposure assessments, measurement calibration and design standards for elevated structures. Full article

Immersive technologies such as virtual reality (VR), augmented reality (AR), mixed reality (MR), and multisensory interaction are increasingly deployed to support the transmission and presentation of intangible cultural heritage (ICH), particularly within tourism and heritage interpretation contexts. In cultural tourism, ICH is often encountered through museums, heritage sites, festivals, and digitally mediated experiences rather than through sustained community-based transmission, raising important challenges for interaction design, accessibility, and cultural representation. This study presents a narrative review of immersive human–computer interaction (HCI) research in the context ICH, with a particular focus on tourism-facing applications. An initial dataset of 145 records was identified through a structured search of major academic databases from their inception to 2024. Following staged screening based on relevance, publication type, and temporal criteria, 97 empirical or technical studies published after 2020 were included in the final analysis. The review synthesises how immersive technologies are applied across seven ICH domains and examines their deployment in key tourism-related settings, including museum interpretation, heritage sites, and sustainable cultural tourism experiences. The findings reveal persistent tensions between technological innovation, cultural authenticity, and user engagement, challenges that are especially pronounced in tourism context. The review also maps the dominant methodological approaches, including user-centred design, participatory frameworks, and mixed-method strategies. By integrating structured screening with narrative synthesis, the review highlights fragmentation in the field, uneven methodological rigour, and gaps in both cultural adaptability and long-term sustainability, and outlines future directions for culturally responsive and inclusive immersive HCI research in ICH tourism. Full article

Background: Clinical practice suggests that objective assessment tools are needed to assess adults with inattention or hyperactivity, informed by the underlying pathophysiology of attention-deficit and hyperactivity disorder (ADHD). This systematic review comprehensively evaluates the current objective assessment methods as an adjunct diagnostic tool for these adults. Methods: We conducted a systematic review of studies investigating various objective diagnostic methods to assess adults with ADHD and healthy controls. The database search occurred from its inception to 23 December 2024. Results: Our search yielded 46 studies that reported on various objective methods to assess adults with ADHD. The MOXO-distracted Continuous Performance Test (MOXO-d-CPT), eye-tracker with MOXO-d CPT, Conners’ Continuous Performance Test—3rd edition (CCPT-3), and oculomotricity can differentiate between true and feigned ADHD or other diagnostic possibilities. The Quantified Behavior Test (Qb Test+) can detect hyperactivity and differentiate it from other psychiatric disorders. Mono-d, CCPT-3, Qb Test+, Test of Variables and Attention (TOVA), Integrated Visual and Auditory Continuous Performance Test (IVA-CPT), and oculomotricity can monitor pharmacotherapy response. Functional near-infrared spectroscopy (fNIRS) offers more promise than structural imaging and demonstrates a moderate level of sensitivity and specificity to differentiate adults with and without ADHD by performing the verbal fluency test. Notwithstanding, electroencephalography (EEG)/event-related potential (ERP) shows potential in diagnosis and treatment monitoring (e.g., neurofeedback training). In addition, transcriptome-based biomarkers have also been explored as diagnostic tools. Conclusion: The diagnosis and monitoring of ADHD in adults come with a unique set of challenges due to psychiatric comorbidity, including depression and anxiety; fluctuation of symptoms over time; and lack of consensus among clinicians and professional organizations to adopt objective tests in the diagnostic process. Our findings support the notion that a combination of clinical assessment and objective biomarkers targeting distinct pathophysiological aspects may enhance the accuracy of ADHD diagnosis. Full article