Search Results (6,363)

Creativity is widely recognized as one of the most important, key competencies supporting the achievement of sustainable development goals. Our paper presents the results of research on the declared level of creativity competence of students at universities located in Kraków (Southern Poland) preparing for the teaching profession. The survey, based on an original questionnaire (49 questions, 7 sections), covered 406 people. The scale was based on an analysis of self-perception of creativity competence among pre-service teachers. Analysis of the results showed that 12.8% of respondents had a high level of creativity, 56.4% had an average level, and 30.8% had a low level. No significant correlations were found between the level of creative competence and gender or age, while place of origin showed a slight tendency to differentiate. Students most often declared reflectiveness, openness to learning, and independence in problem solving, while less often confidence in predicting the effects of their own actions and resistance to routine. The results indicate the significant, albeit partially untapped, creative potential of future teachers. They also emphasize the need to introduce activities in teacher education that strengthen self-confidence, flexibility, and perseverance—qualities necessary to support innovation in education that promotes sustainable development. Full article

As vehicle-centric perception struggles with occlusion and dense traffic, vehicle-infrastructure cooperative perception (VICP) offers a viable route to extend sensing coverage and robustness. This study proposes a feature-level VICP framework that fuses vehicle- and roadside-derived visual features via V2X communication. The model integrates four components: regional feature reconstruction (RFR) for transferring region-specific roadside cues, context-driven channel attention (CDCA) for channel recalibration, uncertainty-weighted fusion (UWF) for confidence-guided weighting, and point sampling voxel fusion (PSVF) for efficient alignment. Evaluated on the DAIR-V2X-C benchmark, our method consistently outperforms state-of-the-art feature-level fusion baselines, achieving improved ${AP}_{3D}$ and ${AP}_{BEV}$ (reported settings: 16.31% and 21.49%, respectively). Ablations show RFR provides the largest single-module gain +3.27% ${AP}_{3D}$ and +3.85% ${AP}_{BEV}$, UWF yields substantial robustness gains, and CDCA offers modest calibration benefits. The framework enhances occlusion handling and cross-view detection while reducing dependence on explicit camera calibration, supporting more generalizable cooperative perception. Full article

This research examines residents’ attitudes toward tourism on Croatian islands as a lens for assessing the sustainability of tourism-led development. In regions where tourism represents a primary economic driver and a major source of local development funding, such reliance has increasingly been recognized as a source of vulnerability. By analyzing survey data collected across islands that differ in geographic size and local characteristics, the research offers an in-depth understanding of how local communities perceive the impacts and future of tourism. Data were analyzed using explanatory factor analysis and cluster analysis. The findings reveal three distinct groups of residents, based on their perceptions of tourism development and shaped by three underlying attitudinal dimensions. The first group strongly supports tourism development. While moderately supportive of tourism as a development strategy, the second group raises concerns about its excessive pressure on local communities and advocates for more stringent regulation. The third group emphasizes the risks associated with tourism-driven growth. These findings underscore the heterogeneity of local perspectives, reflecting varying levels of perceived resilience and vulnerability. The results suggest that areas where residents express greater confidence in tourism and its benefits may exhibit stronger resilience to external shocks and be better suited for tourism-focused development. Conversely, areas where skepticism or concern about tourism prevail may be more susceptible to disruptions and would benefit from more diversified development strategies to build long-term resilience. Full article

To evaluate the creep and fatigue fracture lives of structural acrylic spliced components fabricated via bulk polymerization, and to elucidate the associated fracture mechanisms, this study conducted creep and fatigue tests on spliced coupons annealed at 85 °C and 65 °C, as well as base material coupons. The experimental life data were fitted using log-log linear regression models. Based on statistical analysis, a simple yet robust statistical framework was established for life prediction, featuring three design curves: 97.7% survival curves, improved 95% confidence interval lower bounds, and one-sided tolerance curves. Fractographic examination using scanning electron microscopy (SEM) was performed to characterize macroscopic failure modes. The results indicate distinct threshold behavior between stress levels and both creep and fatigue life. The creep threshold stresses are 25 MPa for the base material, 29 MPa for the spliced coupons annealed at 85 °C, and 17 MPa for the spliced coupons annealed at 65 °C. Corresponding fatigue threshold stress amplitudes are 21 MPa, 22 MPa, and 31 MPa, respectively. Failure in the base material is primarily initiated by randomly distributed internal defects, whereas failure in the spliced coupons is mainly caused by defects within the seam or interfacial tearing. Full article

Background: Diabetes mellitus is a well-established risk factor for surgical site infections (SSIs), particularly periprosthetic joint infections (PJIs) following joint arthroplasty. Although strict glycemic control in the early postoperative period is critical, few studies have evaluated glycemic dynamics using continuous glucose monitoring (CGM) in this setting. This study aimed to characterize early postoperative glycemic patterns using CGM in patients with type 2 diabetes mellitus undergoing lower extremity arthroplasty and to identify factors associated with postoperative hyperglycemia. Methods: We retrospectively analyzed 41 patients with type 2 diabetes who underwent total hip or knee arthroplasty. CGM was used to monitor glucose levels continuously for 48 h after surgery. All patients received standard glycemic management based on a sliding-scale insulin protocol. Patients were classified into two groups: normoglycemia (glucose consistently < 200 mg/dL) and hyperglycemia (glucose ≥ 200 mg/dL at least once within 48 h). Univariable and multivariable logistic regression analyses were conducted to identify predictors of postoperative hyperglycemia. Results: Hyperglycemia occurred in 65.9% of all patients. Univariable analysis identified fasting plasma glucose (FPG), mean postoperative glucose, number of antidiabetic medications, and glucose variability as significant predictors (p < 0.05). In multivariable analysis adjusted for HbA1c, glycoalbumin, and glucose variability, FPG [odds ratio (OR): 1.07; 95% confidence interval (CI): 1.01–1.14; p = 0.024], mean glucose (OR: 1.12; 95% CI: 1.02–1.23; p = 0.017), and glucose variability (OR: 1.19; 95% CI: 1.05–1.35; p = 0.008) remained independently associated with hyperglycemia. Conclusions: CGM revealed a high incidence of early postoperative hyperglycemia despite conventional sliding-scale insulin therapy. These findings highlight the limitations of current glycemic protocols and underscore the potential of CGM as a diagnostic tool to guide individualized glucose management. Future studies should evaluate whether CGM-guided interventions can improve surgical outcomes, particularly in reducing SSI risk among high-risk diabetic patients. Full article

To mitigate the impact of photovoltaic (PV) power generation uncertainty on power systems and accurately depict the PV output range, this paper proposes a quantile regression probabilistic prediction model (TCN-QRBiLSTM) integrating a Temporal Convolutional Network (TCN) and Bidirectional Long Short-Term Memory (BiLSTM). First, the historical dataset is divided into three weather scenarios (sunny, cloudy, and rainy) to generate training and test samples under the same weather conditions. Second, a TCN is used to extract local temporal features, and BiLSTM captures the bidirectional temporal dependencies between power and meteorological data. To address the non-differentiable issue of traditional interval prediction quantile loss functions, the Huber norm is introduced as an approximate replacement for the original loss function by constructing a differentiable improved Quantile Regression (QR) model to generate confidence intervals. Finally, Kernel Density Estimation (KDE) is integrated to output probability density prediction results. Taking a distributed PV power station in East China as the research object, using data from July to September 2022 (15 min resolution, 4128 samples), comparative verification with TCN-QRLSTM and QRBiLSTM models shows that under a 90% confidence level, the Prediction Interval Coverage Probability (PICP) of the proposed model under sunny/cloudy/rainy weather reaches 0.9901, 0.9553, 0.9674, respectively, which is 0.56–3.85% higher than that of comparative models; the Percentage Interval Normalized Average Width (PINAW) is 0.1432, 0.1364, 0.1246, respectively, which is 1.35–6.49% lower than that of comparative models; the comprehensive interval evaluation index (I) is the smallest; and the Bayesian Information Criterion (BIC) is the lowest under all three weather conditions. The results demonstrate that the model can effectively quantify and mitigate PV power generation uncertainty, verifying its reliability and superiority in short-term PV power probabilistic prediction, and it has practical significance for ensuring the safe and economical operation of power grids with high PV penetration. Full article

Legionella is an environmental bacterium capable of causing severe respiratory infections, with outbreaks posing significant public health challenges in developed countries. Understanding public awareness of Legionella transmission, risk perception, and preventive behaviors is crucial for reducing exposure and guiding health education strategies. This study aimed to evaluate the Portuguese population’s knowledge of Legionella infections and their readiness to adopt preventive measures. A structured questionnaire was developed and administered to 239 participants aged 18–76 years across Portugal, collecting socio-demographic data and assessing literacy through statements organized into domains related to Legionella risk, control measures, and public health impact. The results indicate that participants possess moderate to high awareness of Legionella severity, transmission routes, and preventive strategies, yet gaps remain in understanding key risk factors, optimal water system maintenance, and the influence of temperature on bacterial growth. Age, educational attainment, and occupational status were associated with differences in self-assessed literacy levels. Artificial neural network models were applied to classify literacy levels, achieving a near 90% accuracy and demonstrating higher confidence in low and moderate categories. These findings provide insights for designing tailored educational programs, improving public health communication, and enhancing preventive behaviors to reduce Legionella infection risks. Full article

Background: Adolescence is a critical developmental stage marked by vulnerability to psychological difficulties. While family relationships, peer bonds, prosocial behaviors, and health-risk factors have been linked to adolescent mental health, few studies have examined their joint effects in clinical pediatric populations. This study assessed demographic, clinical, relational, and behavioral predictors of psychological difficulties in Italian adolescents. Methods: A cross-sectional sample of 177 adolescents (aged 11–14 years) from a pediatric clinic completed the Strengths and Difficulties Questionnaire (SDQ). The Total Difficulties (SDQ TD) score was the main outcome. Associations were tested with ordinary least squares (OLS) and confirmed using robust MM regression. Bootstrap confidence intervals and Benjamini–Hochberg corrections were applied. Sensitivity analyses excluded the Peer Problems subscale to address part–whole overlap. Results: Higher friendship satisfaction was consistently associated with fewer psychological difficulties, confirming its role as a strong protective factor. Prosocial behavior and male sex were also linked to fewer difficulties in initial analyses, though these associations were less stable after correction. Sensitivity analyses further supported the protective value of friendship satisfaction, even when accounting for overlap with peer problems. Despite relatively low overall levels of psychological difficulties, nearly one-quarter of adolescents met the clinical cut-off for eating disorder risk. Conclusions: Friendship satisfaction was the strongest protective factor, while prosocial behavior and sex showed weaker consistency. Findings suggest that distinct aspects of peer relationships jointly shape adolescents’ psychological outcomes. Interventions promoting social functioning may support mental health in clinical youth populations. Full article

Background: The rising prevalence of neurodegenerative conditions such as dementia underscores the impact of population aging. Consequently, long-term care needs have increased and are often met by family members through informal caregiving, thereby supporting formal care systems by reducing associated costs. These caregivers face physical and mental health challenges, raising concerns about their psychological well-being and prompting interest in both clinical and psychosocial research. Ryff’s eudaimonic model offers a robust framework for the assessment of psychological well-being; yet, in Romania, data on this population segment remain limited. Objective: This study aimed to compare the psychological well-being of Romanian dementia family caregivers with a reference population from the Romanian adaptation of the 54-item Ryff Psychological Well-Being Scale, and to explore how sociodemographic characteristics relate to relevant differences across well-being dimensions. Methods: A cross-sectional study was conducted among 70 Romanian family caregivers recruited from a single clinical hospital in Bucharest, Romania. Caregivers completed the 54-item Ryff Scale (Romanian adaptation), and scores were compared to reference values using one-sample t-tests with bootstrap confidence intervals. The most relevant dimension (purpose in life) was dichotomized and further examined in relation to sociodemographic and caregiving variables using Chi-squared and Fisher’s exact tests. Results: Caregivers reported significantly lower scores compared to the reference population in purpose in life (p < 0.001, d = −1.01), personal growth (p < 0.001, d = −0.91), and positive relations (p = 0.01, d = −0.30). The most pronounced deficit was observed in purpose in life, with 85.7% of caregivers scoring below the reference mean. This dimension was further examined in relation to caregiver characteristics. Retirement status showed a statistically significant association with Purpose in Life, with retired caregivers more likely to report lower scores (χ² (1) = 4.04, p = 0.04), supported by the likelihood ratio test (p = 0.01) and a linear trend (p = 0.05). Additional marginal associations were found for household income (p = 0.14) and whether the patient slept in a separate room (p = 0.15), suggesting possible links between caregiver well-being and economic or environmental conditions. Conclusions: The study findings highlight notable psychological vulnerabilities among Romanian dementia caregivers, particularly in purpose in life and personal growth. Associations with structural and contextual factors such as retirement status, income, and caregiving environment suggest that caregiver well-being is shaped by broader socioeconomic conditions. While the magnitude of these deficits may be underestimated due to elevated stress levels in the reference group, the findings underscore the need for targeted clinical, social, and policy-level interventions aimed at strengthening existential meaning and personal development in culturally specific settings. Full article

Background: Chest radiography (CXR) remains the most common first-line imaging for thoracic abnormalities. While aortic knob width can reflect aortic dilation, no standardized, widely recognized thresholds of clinical utility exist. Methods: This pilot retrospective study analyzed 240 emergency department patients (median age 67 years, 61% male) who underwent both PA CXR and chest computed tomography angiography (CTA) within 7 days. Three aortic knob dimensions (horizontal, oblique, vertical) were measured on CXR and compared with CTA measurements at two anatomical levels: proximal to the brachiocephalic trunk (P-BCT) and distal to the left subclavian artery (D-LSA). Results: The horizontal aortic knob width was most closely related to CTA measurements of P-BCT and D-LSA. A regression model incorporating horizontal knob diameter, age, and sex was characterized with an AUC of 0.884 (95% CI 0.825–0.944) for detecting aortic dilation (>40 mm). Using a conservative threshold with the upper 95% prediction bound exceeding 40 mm led to 100% sensitivity and 54% specificity, with a negative predictive value of 1.00. Conclusions: Simple quantitative CXR measurements of aortic knob width (horizontal), combined with age and sex, can provide additional confidence for excluding aortic arch dilation. Given further validation in diverse populations, if the high negative predictive value of this approach will be confirmed, it may represent a valuable screening tool to guide decisions for advanced imaging, especially due to low cost and wide availability. Full article

Background and Objectives: Diabetic foot gangrene remains a major cause of lower limb amputation, driven by vascular, neuropathic, and infectious mechanisms. Identifying predictors for amputation type is essential to optimizing outcomes and reducing disability. We aimed to analyze the burden of diabetic foot gangrene and the patients’ characteristics according to the type of surgery, minor or major amputations. Methods: We conducted a retrospective observational study including 295 diabetic patients who underwent surgery for foot lesions at a Romanian tertiary center (January 2023–December 2024). Patients were classified according to surgical outcome as minor (toe/foot-level) or major (below/above-knee) amputations. Clinical, demographic, and pathological variables were compared between groups. Statistical analysis was performed with IBM SPSS Statistics 20.0. Categorical variables were expressed as frequencies and percentages, and continuous variables as mean ± SD or median (min–max). Group comparisons used Student’s t-test, Mann–Whitney U, Chi-square, or Fisher’s exact test, and binary logistic regression was applied to calculate odds ratios (OR) with 95% confidence intervals (CI). Results: Among the patients included (mean age 64.8 ± 10.8 years; 69.2% male), 191 (64.7%) underwent minor amputations/debridement and 104 (35.3%) required major amputations. Patients with major amputations were older (66.8 ± 11.3 vs. 63.7 ± 10.4 years, p = 0.012) and less frequently male (56.7% vs. 75.9%, p = 0.001). Lesion extension to the foot or beyond strongly predicted major amputation (p < 0.001). Peripheral arterial disease was more prevalent in the major group (85.6% vs. 65.4%, OR = 3.13, 95% CI = 1.68–5.84), while neuropathy was associated with minor procedures (12.6% vs. 3.8%, p = 0.015). Anemia (70.2% vs. 56.5%, p = 0.021) and leukocytosis (68.3% vs. 49.2%, p = 0.002) were also independent predictors of major amputation. Conclusions: The study highlights the need for early detection, coordinated multidisciplinary care, and personalized assessment of diabetes burden and its complications to minimize the risk of major limb amputation. Full article

This study aimed to analyze which work-related factors may influence the quality of life (QOL) and quality of work life (QWL) of academic teachers from different fields of knowledge, as well as to verify the correlation between QOL and QWL. It is a cross-sectional study in which data were collected using a sociodemographic questionnaire containing work-related questions, the WHOQOL-BREF, and the TQWL-42 instruments. The sample consisted of 284 academic teachers from various disciplines. The total population at the higher education institution (HEI) comprised 386 faculty members, and the sample size was determined using OpenEpi^®, with a 95% confidence level. The results showed no significant differences in QOL and QWL between the different fields of knowledge. However, both QOL and QWL were influenced by several work-related factors, including higher remuneration, holding a statutory employment position, not needing to relocate from one’s home city to work as a professor, adequate lighting, comfortable room temperature, lower noise levels, sufficient material resources, and smaller class sizes. Additionally, a positive correlation between QOL and QWL was observed. In conclusion, both QOL and QWL are influenced by organizational and work-related conditions associated with the academic profession, rather than by disciplinary areas. These findings suggest that the work environment and personal life of academic staff are interdependent, and efforts to improve one may positively impact the other. Full article

In image classification tasks, models may occasionally produce incorrect predictions, which can lead to severe consequences in safety-critical applications. For instance, if a model mistakenly classifies a red traffic light as green, it could result in a traffic accident. Therefore, it is essential to assess the confidence level associated with each prediction. Predictions accompanied by high confidence scores are generally more reliable and can serve as a basis for informed decision-making. To address this, the present paper extends the block-scaling approach—originally developed for estimating classifier accuracy on unlabeled datasets—to compute confidence scores for individual samples in image classification. The proposed method, termed block masking confidence (BMC), applies a sliding mask filled with random noise to occlude localized regions of the input image. Each masked variant is classified, and predictions are aggregated across all variants. The final class is selected via majority voting, and a confidence score is derived based on prediction consistency. To evaluate the effectiveness of BMC, we conducted experiments comparing it against Monte Carlo (MC) dropout and a vanilla baseline across image datasets of varying sizes and distortion levels. While BMC does not consistently outperform the baselines under standard (in-distribution) conditions, it shows clear advantages on distorted and out-of-distribution (OOD) samples. Specifically, on the level-3 distorted iNaturalist 2018 dataset, BMC achieves a median expected calibration error (ECE) of 0.135, compared to 0.345 for MC dropout and 0.264 for the vanilla approach. On the level-3 distorted Places365 dataset, BMC yields an ECE of 0.173, outperforming MC dropout (0.290) and vanilla (0.201). For OOD samples in Places365, BMC achieves a peak entropy of 1.43, higher than the 1.06 observed for both MC dropout and vanilla. Furthermore, combining BMC with MC dropout leads to additional improvements. On distorted Places365, the median ECE is reduced to 0.151, and the peak entropy for OOD samples increases to 1.73. Overall, the proposed BMC method offers a promising framework for uncertainty quantification in image classification, particularly under challenging or distribution-shifted conditions. Full article

Background and objectives: Large language models (LLMs) show promise in automating open-ended evaluation tasks, yet their reliability in rubric-based assessment remains uncertain. Variability in scoring, feedback, and rubric adherence raises concerns about transparency and pedagogical validity in educational contexts. This study introduces CourseEvalAI, a framework designed to enhance consistency and fidelity in rubric-guided evaluation by fine-tuning a general-purpose LLM with authentic university-level instructional content. Methods: The framework employs supervised fine-tuning with Low-Rank Adaptation (LoRA) on rubric-annotated answers and explanations drawn from undergraduate computer science exams. Responses generated by both the base and fine-tuned models were independently evaluated by two human raters and two LLM judges, applying dual-layer rubrics for answers (technical or argumentative) and explanations. Inter-rater reliability was reported as intraclass correlation coefficient (ICC(2,1)), Krippendorff’s α, and quadratic-weighted Cohen’s κ (QWK), and statistical analyses included Welch’s t tests with Holm–Bonferroni correction, Hedges’ g with bootstrap confidence intervals, and Levene’s tests. All responses, scores, feedback, and metadata were stored in a Neo4j graph database for structured exploration. Results: The fine-tuned model consistently outperformed the base version across all rubric dimensions, achieving higher scores for both answers and explanations. After multiple-testing correction, only the Generative Pre-trained Transformer (GPT-4)—judged Technical Answer contrast remains statistically significant; other contrasts show positive trends without passing the adjusted threshold, and no additional significance is claimed for explanation-level results. Variance in scoring decreased, inter-model agreement increased, and evaluator feedback for fine-tuned outputs contained fewer vague or critical remarks, indicating stronger rubric alignment and greater pedagogical coherence. Inter-rater reliability analyses indicated moderate human–human agreement and weaker alignment of LLM judges to the human mean. Originality: CourseEvalAI integrates rubric-guided fine-tuning, dual-layer evaluation, and graph-based storage into a unified framework. This combination provides a replicable and interpretable methodology that enhances the consistency, transparency, and pedagogical value of LLM-based evaluators in higher education and beyond. Full article

Objectives: This study aimed to investigate the diagnostic performance of imaging-based biomarkers from computed tomography (CT) and magnetic resonance imaging (MRI) for prediction of malignant and borderline malignant ovarian tumours. Methods: 195 consecutive patients with suspected primary epithelial ovarian cancer were included from the retrospective “Prognostic and Diagnostic Added Value of Medical Imaging in Staging and Treatment Planning of Gynaecological Cancer” (PRODIGYN) study. The radiological stage, according to the International Federation of Gynaecology and Obstetrics system (rFIGO), magnetic resonance imaging (MRI)-based Ovarian-Adnexal Reporting and Data System (O-RADS-MRI) score, and the mean apparent diffusion coefficient (ADC_mean) were investigated for prediction of ovarian malignancy, with histopathology as reference. The same imaging biomarkers were applied to the borderline tumour cohort (n = 33) to predict malignant/adverse features, such as micro-invasion. Results: The rFIGO stage demonstrated high accuracy for ovarian malignancy, with an area under the curve (AUC) of 0.98 (95% confidence interval (CI) = 0.97–0.99). On lesion level, the sensitivity and specificity of the O-RADS-MRI score to predict ovarian malignancy, after adjusting for correlated data structure, was 1 (CI: 0.96–1) and 0.82 (CI: 0.70–0.90), respectively. The performance of ADC_mean to predict ovarian malignancy on lesion level was moderately high, with AUC = 0.78 (95% CI 0.68, 0.88). Discrimination of adverse features in borderline tumours was not improved. Conclusions: rFIGO and O-RADS-MRI showed excellent performance and outperformed ADC_mean as predictive tools for ovarian malignancy but could not predict adverse features in borderline tumours. Full article