Search Results (84)

Artificial Intelligence (AI) adoption has emerged as a critical priority for governments globally, driven by its transformative potential in improving public service delivery, governance efficiency, and innovation ecosystems. Despite this, substantial disparities exist in AI readiness and adoption levels across countries, necessitating an in-depth exploration of the factors influencing AI adoption. This study leverages data from the Oxford Insights Government AI Readiness Index to model cross-country patterns of government AI readiness through clustering, regression, classification, and explainable machine learning. A Random Forest regression model was first used to estimate the 2024 AI Government Readiness score using lagged 2023 indicators. However, because the dependent variable is a composite index constructed from conceptually related dimensions, this model is interpreted as a lagged score-approximation and benchmarking exercise rather than as an independent forecasting model. The main analytical contribution lies in the clustering-classification framework, which identifies four country-level AI readiness profiles and evaluates the indicators that most strongly distinguish countries across low, moderate-low, intermediate, and high readiness groups. SHAP and permutation-based interpretation methods are used to examine feature contributions, while recognizing that these results indicate model contribution rather than causal effects. The findings underscore the multifaceted nature of AI readiness, emphasizing the interplay between governance, digital infrastructure, and technological investment. Full article

Background/Objectives: Pediatric IgA nephropathy (IgAN) is often considered to have a favorable early course. However, its progression is variable, and the prognostic value of histopathological classifications, such as MEST-C, remains incompletely defined in children. This study aimed to characterize clinicopathological features and the early disease course in pediatric IgAN and to descriptively examine histopathological findings and clinical outcomes. Methods: This retrospective, single-center study included children with biopsy-confirmed IgAN diagnosed between 2016 and 2025. Clinical, laboratory, and histopathological data were collected, and biopsies were assessed using the Oxford MEST-C classification. Follow-up data, including estimated glomerular filtration rate (eGFR), were analyzed descriptively, with follow-up extending from diagnosis to early 2026. Results: Fourteen patients were included, showing heterogeneous clinical presentations. Mesangial hypercellularity was observed in all cases (100%), with frequent endocapillary hypercellularity (78.6%) and segmental sclerosis (57.1%), consistent with a predominance of active lesions. Over a median follow-up of approximately five years, renal function remained stable in 57.1% of patients, declined in 21.4%, and improved in 14.3%, indicating variability in renal function during follow-up and potential reversibility in a subset of patients. One patient (7.1%) developed severe acute kidney injury requiring temporary dialysis, followed by full recovery. Given the descriptive design and limited sample size, no conclusions regarding associations between histopathological findings and renal outcomes can be drawn. Conclusions: Within this small cohort, pediatric IgAN showed variable renal function courses ranging from stability to decline or partial recovery. These findings should be considered descriptive and hypothesis-generating, supporting longitudinal monitoring in larger pediatric cohorts. Full article

Background: High tibial osteotomy (HTO) is an established joint-preserving treatment for medial knee osteoarthritis, but total knee arthroplasty (TKA) may be required if the disease progresses. This pilot study evaluates the radiological and functional outcomes of simultaneous limb axis correction using reverse HTO (rHTO) during TKA compared to a primary TKA control group without previous HTO. Methods: In this retrospective matched-pair study, nine patients with previous valgus HTO underwent varus correction via rHTO combined with TKA using cruciate-retaining implants. Outcomes were compared to a primary TKA control group. Radiographic parameters (MPTA, mLDFA, aHKA, JLO) and clinical scores (Oxford Knee Score, LEFS, KOOS, TAS, FIPS) were assessed pre- and postoperatively. Statistical significance was set at p < 0.05. Results: Preoperative alignment differed significantly between groups (MPTA, aHKA, JLO; all p ≤ 0.001). Postoperatively, both cohorts achieved neutral mechanical alignment showing no statistically significant differences in descriptive parameters. CPAK classification showed convergence to neutral alignment types. Functionally, mean postoperative clinical scores, including KOOS subdomains, LEFS, and Oxford Knee Score, were numerically close between groups, though confidence intervals were wide due to the limited sample size (all p > 0.05). Conclusions: This pilot study demonstrates the technical feasibility of single-stage rHTO and TKA, showing encouraging descriptive clinical and radiographic profiles in this small, highly specific cohort. Given the exploratory nature of this study, larger trials are required to definitively evaluate potential clinical differences. Full article

Rapid and accurate diagnosis of infectious diseases in aquaculture is essential for preventing major economic and ecological losses. Traditional culture-based methods focus on isolation of individual pathogens, and often are burdened with extended processing times, particularly during investigations of polymicrobial infections. Application of Oxford Nanopore Technologies (ONT) sequencing offers a rapid, culture-independent workflow for the identification of bacterial and fungal pathogens directly from fish tissues. Swab and organ samples from four cases (1: Salmo spp.; 2: Cyprinus carpio; 3: Salvelinus fontinalis; 4: Heniochus acuminatus) were analyzed using ONT long-read sequencing for metagenomic screening and bioinformatic classification. The results revealed case-, species-, and tissue-specific microbial profiles, with external tissues showing higher microbial diversity and internal organs enriched in pathogenic taxa. Dominant pathogens included Streptococcus iniae, Aeromonas hydrophila, Pseudomonas spp., and Saprolegnia parasitica, alongside opportunistic zoonotic bacteria such as Escherichia coli and Acinetobacter baumannii. We demonstrate the potential for diagnostic application of ONT sequencing in investigations and detection of multi-pathogen infections, including assessments of microbial community structure changes during disease outbreaks in aquatic species. The presented workflow enables rapid, cost-effective, and comprehensive pathogen profiling, supporting early disease surveillance and improved management in aquatic veterinary practice. Full article

Background: IgA nephropathy (IgAN) is the most common primary glomerulonephritis, characterized by immune dysregulation and progressive renal injury. Among immunoregulatory mediators, interleukin-6 (IL-6) and interleukin-27 (IL-27) have been implicated in inflammatory and immune processes; however, their combined role in IgAN remains poorly understood. Methods: In this study, serum levels of IL-6 and IL-27 were measured in 51 patients with biopsy-proven IgAN and 62 healthy controls using enzyme-linked immunosorbent assay (ELISA). Associations with clinical and histopathological parameters, including the Oxford MEST-C classification, were evaluated. Results: Serum IL-27 levels were significantly higher in patients with IgAN compared to controls (p = 0.001), while IL-6 levels did not differ significantly between groups (p = 0.820). A positive correlation between IL-27 and IL-6 levels was observed in the patient group (Spearman’s rho = 0.287, p = 0.044). Higher IL-27 concentrations were associated with the absence of mesangial proliferation (p = 0.021) and erythrocyturia (p < 0.001), and showed a trend toward lower proportions of global and segmental glomerulosclerosis. ROC analysis demonstrated moderate discriminatory ability for IL-27 (AUC = 0.708), while the combined IL-6/IL-27 model showed only modest improvement (AUC = 0.661). Conclusions: Our findings suggest that IL-27 may play a modulatory and potentially protective role in IgAN, possibly through attenuation of IL-6–mediated inflammatory pathways. Full article

Background: Early external validation studies demonstrated the robust and consistent predictive performance of the International IgA Nephropathy Prediction Tool (IIgAN-PT) across diverse ethnic populations. However, emerging evidence suggests that, in contemporary cohorts of patients with IgA nephropathy, the IIgAN-PT increasingly tends to overestimate the risk of adverse renal outcomes. Subclassification of segmental glomerulosclerosis (S lesions) in the Oxford Classification system (MEST-C) could identify high-risk IgAN patients, with evidence that different S subclassifications respond differently to treatment. Our study aimed to evaluate the predictive performance of the IIgAN-PT in a contemporary Chinese external validation cohort and to optimize its prognostic accuracy by incorporating the most severe and prevalent pathological subclassification of S lesions, NOS⁺Adh⁺, into the original model. Methods: A total of 746 Chinese patients were included with biopsy-proven IgAN in this study. Major adverse kidney events (MAKEs) were defined as death from any cause, initiation of renal replacement therapy, or a 50% decline in eGFR. This study evaluated the discrimination and model fit of three predictive models. The performance of the original and modified IIgAN-PT models was compared and evaluated through reclassification, survival analysis, calibration, decision curve analyses and subgroup analyses. Results: In the study cohort, the median follow-up duration was 4.2 years, during which 77 patients experienced MAKEs. The discriminative ability of the three original models was relatively limited. In contrast, the modified IIgAN-PT incorporating the NOS⁺Adh⁺ subtype of S subclassification demonstrated improved global performance for predicting 5-year risk, achieving a C-index of 0.808 (95% CI, 0.756–0.861). Kaplan–Meier survival curves showed clear risk stratification, particularly between low- and intermediate-risk categories. Reclassification analyses (continuous NRI and IDI) and decision curve analysis further supported enhanced predictive performance, while calibration curves corrected the original model’s risk overestimation. The modified model maintained stable performance across clinically relevant subgroups, including patients with hypertension, proteinuria, or receiving immunosuppression. Conclusions: This study further confirms the independent and clinically relevant prognostic value of the S pathological subclassification. The modified IIgAN-PT model, incorporating the NOS⁺Adh⁺ subtype of S subclassification, demonstrated consistent performance in individualized risk assessment for patients with IgA nephropathy. Full article

Evaluating the trustworthiness of black-box machine learning models remains a significant methodological challenge. Their lack of transparency and interpretability limits applicability, because stakeholders often seek transparency before trusting the results of black-box machine learning models. Explainable AI (XAI) methods provide for human-understandable justifications and informed decision-making of these black-box architectures. Therefore, it is imperative to select the proper XAI model tailored to specific tasks. In this research, we focus on examining four XAI techniques: PEEK, LRP, GRAD-CAM, and LIME to understand how they perform against each other for image classification tasks. We evaluate the performance, robustness, generalizability, noise stability, and computational efficiency of these methods using a globally recognized dataset. With 7390 images, the Oxford IIT pet dataset provides a comprehensive resource for training a custom Convolutional Neural Network (CNN) and VGG16, enabling a consistent evaluation of each XAI method. First, we analyze the saliency maps of the input images and observe the regions predicted by these XAI methods, and then leverage a noise analysis approach to evaluate their performance in terms of accuracy. We further explore the robustness, run-time, and “faithfulness” metrics of each XAI method. In general, we find that these methods can identify a set of input-data features that are critical for accurate classification but also intuitive, such as the outline, face, and eyes of subjects. However, our analysis reveals only marginal consensus among XAI methods in identifying those critical features. Grad-CAM demonstrates strong robustness and stability in VGG16, but the performance on the shallow CNN model remained inconsistent. Full article

Background: Orthodontic camouflage in Class III patients is a widely used treatment approach. However, its application is subject to specific clinical indications. This narrative review aims to synthesize the current scientific evidence and provide an overview of the available methods, their advantages and limitations, in order to guide the appropriate management of Class III camouflage cases. Methods: A literature search was carried out using five main scientific databases without restrictions. Inclusion criteria were all types of articles published on various orthodontic camouflage techniques for Class III malocclusion in adult patients. The string searches included “camouflage” and “Class III malocclusion”. In addition, a manual search was performed to identify further relevant articles. The methodological quality was assessed using the Oxford Centre for Evidence-Based Medicine (OCEBM) classification. Results: This narrative review synthesizes 128 studies on orthodontic camouflage in adult Class III malocclusion. Among the 128 articles included, 110 (86%) were case reports or small case series, corresponding to level 4–5 evidence. The remaining studies were observational in design, most of them retrospective, corresponding to level 2–3 evidence. Extractive or non-extractive treatment can be used for non-surgical treatment of a Class III in adults. In recent years, aesthetic techniques, such as clear aligners or lingual fixed appliances, have been efficiently performed. Carriere Motion III is a fast and efficient method to mask a Class III occlusal relationship. Moreover, the use of temporary anchored devices (TADs) has proven to be an effective and minimally invasive method for managing mandibular distalization and a retraction of the lower incisors. Conclusions: Although most reports are case-based, recent advances such as TADs and clear aligners offer effective non-surgical alternatives for selected mild-to-moderate cases. Careful patient selection remains critical. The evidence was low-quality, and further prospective studies are needed. Full article

Background/Objectives: Acinetobacter baumannii is a critical opportunistic pathogen causing severe healthcare-associated infections, particularly in intensive care units. The global dissemination of carbapenem-resistant A. baumannii (CRAB) and its environmental persistence necessitate continuous genomic surveillance to monitor high-risk clones. Methods: We conducted whole-genome sequencing (WGS), core genome multi-locus sequence typing (cgMLST), and phylogenomic analyses on 26 CRAB isolates collected at the National Institute for Infectious Diseases (INMI) “Lazzaro Spallanzani” IRCCS (September 2023–September 2024). Antimicrobial resistance determinants, virulence-related genes, and capsular (KL) and lipooligosaccharide outer core (OCL) loci were characterized by interrogation of comprehensive bioinformatic pipelines. Results: All CRAB isolates displayed an extensively drug-resistant (XDR) phenotype, with a shared resistance pattern to carbapenems, aminoglycosides, fluoroquinolones, fosfomycin, and sulfonamides, while being susceptible only to colistin and cefiderocol. The carbapenemase gene bla_OXA-23 was detected in all CRAB isolates, together with clone-specific bla_OXA-51-like variants. For all isolates, the resistome profile fully matched the observed resistance phenotype. All isolates belonged to the International Clonal Lineage II (ICL II), Pasteur Sequence Type (ST) 2, and Oxford ST369, ST208, and ST455. Integration of cgMLST data with phylogenomic analyses and genome-based classification of KL and OCL loci revealed five distinct clusters, each one including nearly identical isolates, indicating both intra-hospital dissemination and possible inter-hospital transmission. Virulome profiling revealed heterogeneous repertoires of virulence-associated genes, resulting in cluster-specific patterns, while patristic analysis identified phylogenetic clusters linking the study isolates to other Italian and other European lineages. Conclusions: This study underscores the complex genomic landscape of CRAB in our setting, driven by the circulation of different ICL II clonal types, and reinforces the urgency of integrated genomic surveillance and robust antimicrobial stewardship to mitigate the spread of high-risk XDR A. baumannii clones. Full article

IgA nephropathy is the most common primary glomerulonephritis worldwide and a leading cause of chronic kidney disease and kidney failure, with geographic and ancestral variation and a course ranging from asymptomatic urinary abnormalities to progressive loss of kidney function. This narrative review links the multi-hit model to risk stratification, biomarkers, current management, and emerging therapies, and highlights implementation gaps. Risk assessment is longitudinal, prioritizing proteinuria and estimated glomerular filtration rate trajectories and integrating Oxford MEST-C, prediction tools, and biomarker and multi-omics approaches, while recognizing limitations in histologic reproducibility and model calibration. Current management is anchored in optimized supportive care aimed at sustained proteinuria reduction and kidney protection, including intensive blood pressure control with maximal tolerated renin–angiotensin system blockade, dietary sodium restriction and lifestyle measures, and sodium–glucose co-transporter 2 inhibitors for eligible patients. For selected higher-risk patients with persistent proteinuria despite optimized supportive care, immunomodulatory strategies are discussed, including systemic corticosteroids and targeted-release budesonide (Nefecon), emphasizing structured toxicity risk mitigation and cautioning against assuming interchangeability among alternative oral budesonide formulations. Emerging therapies are organized around mechanism-aligned targets across the BAFF/APRIL axis, complement pathways, and endothelin-based approaches, with growing interest in sequencing and combination regimens layered on supportive care. Key gaps include reliance on surrogate endpoints, limited long-term durability and safety data, and uneven evidence for special populations. Full article

Background/Objectives: Stroke is a sudden neurological event caused by disrupted cerebral blood flow and represents a leading cause of acquired disability worldwide. Motor impairments and spasticity are among the most prevalent sequelae, significantly limiting functional independence and quality of life. Non-invasive electrotherapy has emerged as a complementary strategy in neurorehabilitation aimed at enhancing neuroplasticity and improving motor recovery. To systematically review current evidence regarding the effectiveness of non-invasive electrotherapy modalities in the rehabilitation of motor sequelae and spasticity following stroke, and to examine their theoretical and clinical rationale. Methods: A systematic literature review was conducted in accordance with PRISMA 2020 guidelines. The protocol was prospectively registered in the Open Science Framework (OSF). A comprehensive search was performed in Pubmed, Web of Science (WoS), and Scopus for studies published up to 14 November 2023, using the terms “Electric Stimulation Therapy” and “Stroke”. The methodological quality was assessed using the PEDro scale. The levels of evidence were classified according to the Oxford Centre for Evidence-Based Medicine criteria, and the risk of bias was evaluated using the Cochrane Risk of Bias tool (RoB 2). Results: Sixteen studies were included in the review. The analyzed interventions comprised neuromuscular electrical stimulation (NMES), transcutaneous electrical nerve stimulation (TENS), functional electrical stimulation (FES), neuromuscular electrical stimulation combined with transcranial magnetic stimulation (NMES + rTMS), transcranial direct current stimulation (tDCS), and afferent electrical stimulation (AES). Overall, the methodological quality of the included studies ranged from moderate to high, with PEDro scores between 6 and 9 out of 10. According to the Oxford Centre for Evidence-Based Medicine classification, most studies corresponded to level 1b evidence, while a smaller proportion were classified as level 2b. A risk of bias assessment using the Cochrane RoB 2 tool showed that the majority of the included studies presented a low risk of bias across most domains, although some concerns were identified in the domains of randomization and measurement in a limited number of trials. Across modalities, consistency within group improvement in motor function and spasticity was reported. However, between group comparisons with conventional rehabilitation were often inconsistent and did not consistently demonstrate superiority. The variability in stimulation parameters, intervention duration, and outcome measures further limited direct comparisons across studies. Conclusions: Non-invasive electrotherapy appears to be a safe and promising adjunct to conventional stroke rehabilitation. Nevertheless, further high-quality studies are required to clarify the underlying neurophysiological mechanisms and to establish standardized treatment protocols. Full article

Pseudo-multiview learning improves classification by integrating complementary feature representations, but its performance degrades as the number of psuedo-views increases due to model collapse and ineffective feature scaling. This paper introduces a multiscale grid architecture that extracts structured, scale-adaptive features to stabilize evidence aggregation in pseudo-multiview learning. The proposed design enables efficient handling of difficult classification scenarios by enforcing balanced multiscale representation and reducing redundancy across psuedo-views. Extensive experiments on challenging real-world datasets, including BreakHis (40×, 100×, 200×, 400×), Oxford-IIIT Pet, and Chest X-ray, demonstrate consistent gains in accuracy and stability over the original pseudo-multiview framework and other baseline models. The results confirm that grid-based multiscale feature extraction provides a reliable means to enhance pseudo-multiview learning, particularly in settings where prior methods struggled to generalize. Full article

Primary glomerulonephritis encompasses a diverse group of kidney diseases with variable clinical trajectories and outcomes. Accurate prognostic stratification is critical for guiding individualized management and improving long-term renal survival. This narrative review synthesizes current evidence on the prognostic value of histological grading systems, circulating and urinary biomarkers, and integrative risk prediction models across major primary glomerulonephritis subtypes, including IgA nephropathy, membranous nephropathy, and focal segmental glomerulosclerosis. Emphasis is placed on the utility of established classification systems (e.g., Oxford, MEST-C, chronicity scores), emerging tissue and fluid biomarkers (e.g., PLA2R antibodies, complement components, cytokine profiles), and the validation of multivariable prognostic tools and nomograms. We highlight areas of convergence between histopathologic lesions and molecular markers, as well as the evolving role of machine learning in predictive modeling. Ultimately, combining morphological, biochemical, and algorithmic tools holds promise for precision risk assessment and treatment tailoring in primary glomerulonephritis. Full article

Background: This scoping review aimed to identify and synthesize the evidence on mindfulness-based interventions targeting mental health outcomes in working adults. Methods: A comprehensive search was conducted in four electronic databases (PubMed, Scopus, Web of Science, and OTseeker) up to October 2025. The review followed PRISMA-ScR guidelines. Methodological quality was evaluated using the Oxford Centre for Evidence-Based Medicine (OCEBM) classification. The protocol was prospectively registered on the Open Science Framework (OSF). Results: A total of 1803 records were identified, of which nine randomized controlled trials met the inclusion criteria. The included studies examined Mindfulness-Based Stress Reduction, Mindfulness-Based Self-Care, Mindfulness-Oriented Therapy, and digital mindfulness interventions. Overall, mindfulness interventions demonstrated beneficial effects across several mental health domains. Conclusions: Mindfulness-based interventions show promising benefits for improving mental health among working adults. Their structured, brief, and adaptable formats support their feasibility for integration into occupational health programs and workplace mental health promotion. Full article

Background: The relationship between comorbidity burden, as measured by the American Society of Anesthesiologists (ASA) classification, and functional recovery after total hip replacement (THR) remains uncertain. This study aimed to clarify whether ASA grade independently predicts postoperative patient-reported outcomes. Methods: We conducted a retrospective analysis of 218 consecutive patients from a prospectively maintained institutional registry who underwent primary unilateral THR between March 2021 and March 2024 in a single center. Patients were stratified into ASA 1–2 and ASA 3 groups. The Oxford Hip Score (OHS, 0–48) was collected preoperatively and at 1 week, 3 months, and 6 months postoperatively. Between-group differences were assessed, and multivariable linear regression was used to identify predictors of 6-month OHS. Results: Compared with ASA 1–2 patients, ASA 3 patients had lower preoperative OHS and longer hospital stay, but both groups showed substantial improvement over time and achieved excellent mean OHS at 6 months. In the adjusted model, higher ASA grade remained an independent negative predictor of 6-month OHS, whereas higher preoperative OHS and BMI were positive predictors. Conclusions: Despite presenting with worse baseline function and requiring longer hospitalization, ASA 3 patients experienced clinically meaningful recovery and achieved favorable 6-month outcomes after THR. Higher ASA status should therefore inform perioperative optimization rather than preclude surgery. Full article