Search Results (523)

Fat-poor angiomyolipoma (fp-AML), a common benign renal mass, closely mimics renal cell carcinoma (RCC) on preoperative computed tomography (CT), frequently resulting in unnecessary surgical intervention. This multicenter retrospective study aimed to develop and externally validate an AI-assisted radiomics model based on triphasic contrast-enhanced CT to accurately distinguish fp-AML from RCC. A total of 655 eligible patients with sporadic solid renal lesions were enrolled and divided into a training cohort (n = 364), an internal test cohort (n = 156), and an independent external validation cohort (n = 135), with a stable fp-AML to RCC ratio of approximately 1:4. Tumor segmentation was performed using an nnU-Net-assisted workflow with radiologist refinement, followed by cross-phase image registration, radiomic feature extraction, LASSO feature reduction, and random forest classifier construction. The established model achieved favorable and stable diagnostic performance across all cohorts, with AUCs of 0.868 in the training and internal test sets and 0.803 in the external validation set, maintaining reliable discrimination even in the small renal mass subgroup (≤4 cm). This externally validated AI radiomics model demonstrates promising diagnostic performance across centers and may serve as a preoperative decision-support tool for differentiating fp-AML from RCC; however, prospective multicenter validation is required before routine clinical implementation. Full article

Background: Artificial intelligence (AI) systems, including large language models, are increasingly used in clinical practice, whether consulted informally by clinicians or introduced by employers into decision workflows. It remains unclear how physicians attribute moral responsibility when a decision follows an AI recommendation and whether that attribution varies with the type of decision at stake. Methods: We conducted a cross-sectional, within-subject vignette survey of physicians in Romania. Each respondent rated the same three scenarios—urgent clinical, elective clinical, and administrative—in which a physician followed an AI recommendation under two extenuating institutional constraints. Five-point Likert items addressed the mitigation of blame by circumstances, physician responsibility despite the AI recommendation, and institutional co-responsibility. Analyses were non-parametric, with corrections for multiple testing. Results: Among 72 physicians from 17 specialties, respondents endorsed full personal responsibility in every scenario, including the administrative one, with no significant difference between scenarios. They rejected extenuating circumstances as mitigating in both clinical scenarios but were divided about them in the administrative scenario, which had the largest effect. Institutional co-responsibility was endorsed alongside personal responsibility rather than in place of it, and the two attributions were largely uncorrelated. No demographic association survived correction, although the study was not powered to detect small-effect sizes. Conclusions: Physicians treated AI as an instrument rather than a bearer of responsibility, which is unsurprising. The substantive findings lie elsewhere: personal responsibility was retained across all decision contexts, while what varied was the admissibility of institutional constraints as excuses and the emphasis placed on the institution’s share. Because respondents did not treat responsibility as a fixed quantity to be divided, the pattern is consistent with distributed-responsibility accounts rather than with a responsibility gap, though attitudinal data cannot adjudicate between normative accounts. The findings are exploratory and require confirmation in larger, more representative samples. Full article

The rapid evolution of AI-driven language technologies has inadvertently widened the gap between high-resource and marginalised languages. Despite significant progress in AI-driven translation for high-resource languages, low-resource languages remain underrepresented due to limited data, a lack of benchmarks, and evaluation challenges. This study presents a comprehensive systematic review of machine translation for low-resource languages, focusing on advances in neural machine translation (NMT) and large language models (LLMs) between 2017 and 2025. Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, 63 studies were selected from the 1696 articles in the Scopus, Web of Science, and Google Scholar databases. The review identifies five dominant methodological approaches: data augmentation, back-translation, transfer learning, pre-training, and parameter-efficient fine-tuning. The findings reveal that model performance is highly dependent on resource availability: transformer-based NMT excels in moderate data settings, while LLMs demonstrate promising zero-shot and few-shot capabilities in extremely low-resource scenarios. Hybrid NMT–LLM approaches emerge as a particularly effective paradigm. The study also highlights critical challenges, including the absence of standardised benchmarks, over-reliance on inadequate evaluation metrics such as Bilingual Evaluation Understudy (BLEU), limited human evaluation, and significant geographic and linguistic underrepresentation. Additionally, ethical concerns related to bias, cultural representation, and community engagement are increasingly relevant. The findings contribute to advancing inclusive and equitable AI-driven language technologies. Full article

Large bibliographic databases require careful interpretation because publication trends and subject-area counts are shaped by database coverage, document-type selection, and classification practices. This study examines longitudinal changes in articles and reviews indexed in Scopus between 2010 and 2025, focusing on total Scopus-indexed output, the Business, Management and Accounting (BMA) category, and source-derived subject-area classification multiplicity. A longitudinal bibliometric design combined annual publication counts, growth rates, comparisons between the 2010–2021 baseline and the 2022–2025 recent observation window, exploratory segmented trend models, and the Subject-Area Multiplicity Ratio (SAMR). The SAMR was calculated as the annual sum of source-derived Scopus ASJC subject-area counts divided by the number of unique indexed articles and reviews. It is used as a descriptive ratio of classification multiplicity, not as a measure of article-level interdisciplinarity. The results show sustained but heterogeneous growth across subject areas. BMA recorded comparatively stronger growth during 2022–2025, whereas total Scopus-indexed output did not display a generalized discontinuity after 2022. Generative AI is treated only as contextual background, not as an explanatory factor. The SAMR increased from 1.654 in 2010 to 1.819 in 2025, indicating that summed subject-area counts increasingly exceeded unique-document totals. These findings support cautious interpretation of longitudinal bibliometric indicators in research assessment contexts. Full article

Recent statistics reported by the World Health Organization and the International Agency for Research on Cancer indicate that the global incidence of bladder cancer has continued to increase in recent years, particularly in industrialized countries. Therefore, the timely diagnosis of early-stage bladder cancer is of great clinical importance for improving patient prognosis and treatment outcomes. In this context, computational optical sensing frameworks that integrate Spectrum-Aided Visual Enhancer (SAVE) technology with cystoscopy have attracted significant attention to overcome the limitations of conventional visual data interpretation. In this study, an AI-driven optical biosensing framework was evaluated using 1372 white-light cystoscopy (WLC) images of bladder cancer (RGB-WLC) collected in collaboration with Chung Shan Medical University Hospital. Hyperspectral conversion technology was applied to extract precise spectral information from the white-light images. Subsequently, dimensionality reduction was performed based on the characteristic wavelengths of narrow-band imaging cystoscopy at 415 nm and 540 nm to generate hyperspectral reconstructed narrow-band images. The images were categorized into Ta stage (Ta), above T1 stage (Above T1), and four additional classes. The dataset was divided into training and testing sets to establish both a standard white-light cystoscopy model (RGB-WLC) and an advanced hyperspectral biosensing model utilizing the YOLOv8 architecture for enhanced pattern recognition. Model performance was evaluated using sensitivity, F1-score, and overall accuracy. The standard RGB-WLC model achieved an accuracy of 0.852, whereas the SAVE-based biosensing model achieved an accuracy of 0.948, representing an improvement of approximately 11.27%. The results demonstrate that combining algorithmic hyperspectral reconstruction with deep learning architectures effectively addresses the challenges of clinical data interpretation and significantly enhances the detection and staging performance of bladder cancer imaging. Full article

Determining the parameters of a manipulator for optimal performance is a challenging task. This is primarily due to possible conflicting objectives, various tasks that should be considered, and the highly non-linear behavior that is involved. This work proposes an enhanced version of the concept-based design space exploration (C-DSE) approach for the design of manipulators. According to the C-DSE approach, prior to the search, the designers divide the set of feasible solutions into meaningful subsets, which are termed concepts. The design space exploration involves a simultaneous search for optimal solutions within each of the pre-defined concepts. This enhanced framework integrates the following: (1) kinematics, dynamics, and control co-design, and the simultaneous optimization of manipulator morphology and controller parameters; (2) surrogate-assisted optimization using Gaussian process (GP) and neural network (NN) models to reduce computational cost; (3) approximately 30 performance metrics spanning kinematic, dynamic, structural, control, and task performance domains; (4) task-aware feasibility verification applying a multi-level hierarchy; (5) a generative AI integration pathway using diffusion models and LLM-guided concept generation (proposed in this preliminary investigation). The results demonstrate a 95.7% reduction in high-fidelity function evaluations (50,000 to 2150), corresponding to a 23.3 times reduction in evaluation count and a 6.6 times reduction in wall-clock computation time (25 h to 3.8 h). Co-design yields up to a 35% improvement in energy efficiency and a 28% reduction in tracking error compared to sequential morphology-only optimization. Full article

Generative AI (GenAI) adoption in higher education (HE) raises significant ethical concerns. The focus is shifting from rules- or outcomes-based learning environments towards the development of moral character, personality traits, integrity, and practical wisdom (phronesis). However, most existing AI ethics validation instruments are predominantly shaped by Global North paradigms. In Global South HE contexts, in particular, open distance e-learning (ODEL) HE institutions (HEIs) characterised by limited direct supervision and a digital divide, validation remains scant. Ethical risks are intensified by the adoption and integration of GenAI tools, such as large language models (LLMs), to enhance teaching, learning, research, and student support, thus recognising the need to develop and validate virtue ethics scales. The current paper attempts to address this gap by validating the Virtue Ethics Measurement Scale (VEMS) within South Africa’s largest comprehensive ODEL institution. Guided by the positivist paradigm, a 36-item cross-sectional survey of 503 undergraduate and postgraduate students measured six virtue dimensions (justice, honesty, responsibility, care, prudence, and fortitude). Confirmatory factor analysis (CFA) compared four competing models. The single-factor model showed poor fit, rejecting unidimensionality. A second-order hierarchical model demonstrated an acceptable fit (χ²/df = 2.992, CFI = 0.933, RMSEA (Root Mean Square Error of Approximation) = 0.063, SRMR (Standardized Root Mean Squared Residual) = 0.043) with subscale reliabilities ranging from Cronbach’s α = 0.84 to 0.90, supporting a multidimensional yet hierarchical virtue structure. The VEMS offers a psychometrically sound instrument for evaluating ethical AI use in ODEL institutions. This aligns with virtue ethics theory, which emphasises that moral character is a constellation of dispositions (e.g., honesty, care, prudence) rather than a single trait. The VEMS thus enables HEIs to assess students’ virtues, design targeted ethics capacity-development programmes, and inform policy reform for responsible GenAI adoption in under-researched Global South HE settings. Full article

Artificial intelligence (AI) is transforming requirements elicitation: machine learning, natural language processing (NLP), and large language models (LLMs) now identify software requirements automatically from the textual data that surrounds every project—user feedback, specifications, regulations, and stakeholder transcripts. This paper presents a systematic mapping study of 74 peer-reviewed primary studies on AI-based automated requirements elicitation published between 2021 and 2025, identified from five databases following PRISMA 2020 and classified by AI technique, textual source, elicitation activity, and application domain. The evidence is divided into two equally sized source families—user feedback and agile artefacts versus formal documentation—each coupled to the AI techniques that suit its signal profile. Fine-tuned transformer encoders set the performance ceiling and, task-for-task, still outperform far larger generative models, while LLMs extend elicitation to long regulatory documents, multilingual feedback, and structured outputs. The central finding concerns automation depth. AI identifies requirements with consistently high accuracy (routinely F1 0.8 and above), but automation thins at every subsequent step: 51% of approaches structure what they identify, 23% consolidate them, and only 8% engineer stakeholder validation into the loop. This leaves the steps that turn candidates into agreed requirements largely manual. Benchmark fragmentation (77% custom datasets), thin industrial validation (14%), and skewed non-functional coverage compound this gap. The resulting map gives researchers an evidence-derived agenda for deepening automation, and practitioners guidance on which techniques the evidence supports for each elicitation task and textual source. Full article

Background/Objectives: Although efficacious in randomized trials, the real-world adoption of AI-driven dietary management applications remains uncertain across diverse dining contexts and populations. Methods: This field-based observational study was conducted over 18 days at three real-world dining sites in Shanghai, China, enrolling 181 participants stratified into three groups based on food service style and customer attribute. A cross-sectional survey was administered on day 9, followed by a 9-day prospective usage tracking period. Results: After adjusting for sex, Group 2 (staff cafeteria with fixed-portion dishes) had the highest adjusted mean usability score at 71.20 (p < 0.001). Group 3 (community canteen) had the highest mean scores for information quality (16.57, p = 0.03) and perceptions of intended use in nutrition (12.01, p = 0.08). However, Group 3 recorded zero active usage sessions despite favorable initial perceptions. Conclusions: Favorable user perceptions of this AI-driven dietary management tool did not automatically translate into adoption. Scenario-specific usability and digital divide constraints define the boundary of real-world efficacy; moreover, AI may amplify existing dietary self-management behaviors rather than creating them de novo. Full article

Bispecific antibodies (bsAbs) are engineered to recognize either two distinct antigens or two different epitopes on the same antigen within a single molecule. This design varies according to the intended indication and mechanism of action; the factors considered during design are critical determinants of antigen binding, pharmacological activity, productivity, and safety. In this review, bsAbs are classified into fragment-based formats and Fc-containing IgG-like formats, with the latter further divided into symmetric and asymmetric architectures. Based on this structural framework, we discuss how key design parameters—including valency, epitope geometry, affinity and binding kinetics, and linker architecture—influence avidity, immune synapse formation, receptor clustering, signaling modulation, and toxicity profiles. We further compare preclinical and clinical examples across representative target combinations, including CD19 × CD3, CD20 × CD3, BCMA × CD3, HER2 × HER2, and EGFR × MET, to illustrate how different molecular formats yield distinct therapeutic outcomes even when the target combinations are similar. By linking structural classification with mechanism-based interpretation and within-target comparisons, this framework relates individual design variables directly to their preclinical and clinical consequences. Finally, we describe how the energy-based molecular modeling platform Rosetta and the deep-learning-based structure-prediction system AlphaFold are applied to support interface optimization, chain-pairing control, epitope geometry prediction, and structure-guided candidate prioritization. Overall, this review can provide a structure–function framework for bsAb design by integrating key structural determinants, their functional consequences, and emerging AI-based predictive strategies to facilitate the selection of optimal molecular architectures for specific therapeutic applications. Full article

Synthetic data are increasingly proposed as a strategy for addressing data scarcity and representation imbalance in medical AI, particularly for paediatric populations and darker skin tones. However, visually plausible synthetic images may still contain clinically implausible features or fairness-relevant inconsistencies that are not adequately captured by automatic image-quality metrics. In this study, we present and empirically evaluate a clinician-guided framework for validating and selecting synthetic paediatric dermatology images. The framework combines a clinician-facing evaluation platform with structured assessments of visual realism, mask quality, diagnostic plausibility, confidence, and skin-tone relevance. Four clinicians with complementary expertise in paediatrics and dermatology completed 282 assessments of 93 real and synthetic images. Synthetic images were often rated as visually realistic but showed lower inter-rater agreement and weaker mask-quality assessments than real images. Clinician realism and confidence ratings were then used to divide 30 synthetic images into 18 approved and 12 non-approved images. To assess downstream utility, we compared a real-only ResNet50 classifier with classifiers augmented using all synthetic images, clinician-approved synthetic images, or non-approved synthetic images. Across three patient-level experimental splits, the clinician-approved condition achieved the strongest overall classification performance and the largest gains for the under-represented Dark-Skin subgroup. Because the Dark-Skin subgroup contained only seven patients and the synthetic subsets differed in size and disease composition, these fairness results should be interpreted as exploratory. The present study therefore provides evidence for clinician-guided validation and data curation rather than for a completed iterative generator-retraining process. Future work will evaluate whether clinician feedback can also support repeated generative-model refinement in larger, multi-centre datasets. Full article

Our first objective was to quantify the associations between health-related events (HRE) before insemination, the relative increase in estrus intensity (REI) at insemination, and the probability of cow-level pregnancy per artificial insemination (P/AI) in organic Holstein dairy cows. Quantifying these associations may aid on-farm decision-making, such as setting the voluntary waiting period, choice of type of semen, do-not-breed and culling decisions. A second objective was to develop predictive models to estimate P/AI based on readily available data, and present common goodness-of-fit results also used in the machine learning community. All data were collected from a certified organic dairy farm in the western USA from 2019 to 2021. Health-related and reproduction data were obtained through Dairy Records Management Systems (DRMS; Raleigh, NC, USA). Activity data were collected using pedometers (IceRobotics, Stirling, UK) mounted on the rear legs. The REI, defined as walking steps per hour before insemination divided by the cow’s baseline steps per hour, was available for 17,238 inseminations from 4759 cows. The REI was categorized as ≤200%, >200–400%, >400–600%, or >600%. The HRE were available for 65,684 inseminations from 13,365 cows. The HRE were categorized as mastitis, metabolic disease (i.e., hypocalcemia, ketosis, displaced abomasum, digestive problems), reproductive disease (i.e, metritis, endometritis, pyometra, retained fetal membranes), lameness, 2 different diseases, ≥3 different diseases, or as healthy (none of these diseases prior to insemination). Combinations (COMBO) between REI categories and 0, 1, or ≥2 HRE were also created. Data were split into training and test sets. The training data were used to fit three logistic regression models that included either HRE, or REI, or COMBO. Each of the three models also included the covariates of 3-mo herd-average P/AI prior to insemination, days in milk, and the fixed effects of parity, insemination season, days after the previous insemination or days to 1st insemination. A random effect accounted for repeated inseminations within cow. Parameter estimates, odds ratios, and the estimated marginal means of the estimated P/AI of the fixed effects were obtained from the logistic regression models. The models’ estimates were applied to the test datasets, and discrimination and calibration statistics were calculated to judge goodness-of-fit. Unadjusted mean P/AI were 0.31, 0.28 and 0.28 for the HRE, REI and COMBO training datasets. For the HRE model, estimated P/AI ranged from 0.20 (≥3 different HRE) to 0.30 (healthy). The estimated P/AI associated with four REI categories were not different from 0.27 in the REI model. The estimated P/AI associated with the combinations of HRE and REI in the COMBO model varied from 0.18 after ≥2 HRE and >200–400% REI, to 0.30 when inseminations were in healthy cows with REI >600%. Inseminations in older cows, in the spring, and outside 18–24 d after the previous insemination were also associated with lower estimated P/AI. The area underneath the Receiver Operating Characteristic curve ranged from 0.57 (COMBO) to 0.60 (HRE) for the test data, indicating fair discrimination ability of the models. Calibration plots showed that the prediction models produced unbiased predicted P/AI. In conclusion, the results showed no conclusive evidence of greater estimated P/AI related to greater REI as a measure of estrus activity. More HRE were associated with lower estimated P/AI. Combinations of low REI and more HRE were associated with notably decreased estimated P/AI. The logistic regression models produced unbiased predicted P/AI. We found no evidence that the strength of the relationship between REI and P/AI depended on the HRE category. The applications of the results are as follows. First, these predictive models may help inform insemination decisions in organic dairy cows, although further external validation is recommended, and the discriminatory performance is weak. Second, a variety of goodness-of-fit statistics were calculated to allow comparisons of the current logistic regression analyses with future analyses made by other machine learning techniques. Full article

Virtual reality (VR) technology presents the potential of transforming science education. Chemistry education, in particular, has traditionally been limited by the safety, infrastructure, and cost of chemistry laboratories, as well as the inability of traditional methods to visualize microscopic phenomena within chemistry. While there are various virtual reality systems that have been published that describe the capabilities of performing individual functions within the virtual chemistry classroom (such as simulating chemistry experiments or visualizing molecules), few have incorporated the various components of an effective chemistry classroom within their software, and the literature offers no unifying architectural framework or accompanying set of design principles for integrating those components systematically. This paper presents a design framework and reference architecture for an immersive virtual reality smart chemistry laboratory. The framework has not yet been implemented as a software system; its value lies in a principle-driven design blueprint—layers, responsibilities, inter-layer data flows, and design rules—that is intended to guide future implementations. The framework divides the virtual chemistry classroom into five major layers and the data that flows between those layers. Each layer has a relationship with other theories of learning, such as constructivism, experiential learning theory, and cognitive load theory, as well as a rule-based system for providing guidance to the user to enhance their learning of chemistry topics. The intelligence of the framework is deliberately staged: its core is a rule-based, explainable expert system that is implementable with current technology, while AI and generative-AI capabilities are positioned as optional future augmentations of specific layers. Additionally, a model for governing the data collected from the virtual chemistry laboratory is presented. Both existing literature and case studies of third-party virtual reality systems are reviewed to indicate the capabilities of each of these systems; these systems are not the implementations of the current framework. As a result, the virtual reality chemistry classroom can enhance engagement in chemistry concepts and topics, allow for accessibility to chemistry concepts, and reduce the potential for chemical exposure to students. The scalable design framework and reference architecture described in this paper can be used to guide the development of next-generation virtual chemistry classrooms. Future efforts in regard to this technology would focus upon actually implementing the proposed framework, deploying it into chemistry classrooms, and evaluating the learning of students who utilize the learning environment. Full article

Free generative AI (GenAI) services appear to democratize access, but freemium platform models can create post-access stratification after basic connectivity and initial use have been achieved. This article conceptualizes paid commitment to GenAI as an emergent outcome of a socio-technical system composed of platform tiers, user capabilities, task environments, economic constraints, affective feedback, and institutional access. Using a cross-sectional 2025 survey of 2000 South Korean adults aged 18–69, the study models the observed states of non-use, free use, and paid commitment through sequential logit models. Descriptive results show that 70.6% of respondents used GenAI in the previous year, yet only 20.6% of users, or 14.6% of the total sample, paid for access. Practical AI literacy is positively associated with both initial use and paid commitment, whereas critical AI awareness is associated with initial use but not payment. Negative emotions are negatively associated with initial use but positively associated with paid commitment among users, a pattern consistent with vigilant and intensive engagement. The premium digital divide is defined not as a fourth level of the digital divide, but as a tier-mediated extension of second-level inequality that may condition access to later educational, occupational, or civic benefits. Because the data are cross-sectional and combine different GenAI platforms, the estimates are ecosystem-level associations and do not establish causal direction or platform-specific effects. Full article

The compelling visualization of Digital Elevation Models (DEMs) constitutes a vital intersection between Geographic Information Science (GIS) and the digital humanities. Nevertheless, traditional Generative Adversarial Networks (GANs) frequently demonstrate a “geography-blind” characteristic, resulting in structural “topographic drift” by dissociating geomorphic complexity from cartographic constraints. To overcome this limitation, we propose Geo-InkGAN, a geo-heuristic framework that integrates geographic principles with generative processes to achieve high-fidelity ink-wash style synthesis. A key component of our approach is an adaptive optimization strategy grounded in the Slope Standard Deviation (SSD). By establishing a quantitative relationship between geomorphological entropy and the cycle-consistency loss weight (${\lambda }_{cyc}$), we effectively address the Pareto trade-off between geomorphic accuracy and esthetic representation. Our results indicate that alluvial plains benefit from low-intensity constraints to facilitate fluid ink diffusion, whereas rugged terrains require high-intensity constraints to maintain the integrity of the topological framework. Additionally, the HCEG-SE mechanism (Hillshade-Contour Edge-Guided Stroke Enhancement) narrows the semantic divide between terrain skeletons and artistic textures by combining multi-directional non-photorealistic rendering with precise edge extraction techniques. Evaluated across five geomorphologically diverse regions—from karst towers to loess plateaus—Geo-InkGAN demonstrably surpasses existing benchmarks in Geomorphological Structure Correlation (GSC). This geomorphology-aware approach advances the scientific rigor of AI-driven cartography and offers a refined methodology for the cultural representation of digital twin landscapes. Full article