Search Results (1,979)

The rapid development of intelligent systems has created a need for techniques capable of handling complexities and developing in an automatic way. The integration of machine learning in model-driven engineering (MDE) offers several advantages for the development and improvement of complex and intelligent systems. While machine learning (ML) also offers robust techniques and MDE has systematic approaches aimed at code generation and abstraction, in this review, while presenting the principles of MDE and ML, the article also critically explores the integration of ML in MDE. Starting with the fundamental concepts of MDE, then the principles and algorithms of ML, the focus of the discussion is on how machine learning techniques can improve model-driven engineering processes. By presenting the motivations for their combined use in the development of intelligent systems, based on the recent literature, the article describes the challenges and potential future directions, noting that the integration of machine learning into model-driven engineering not only accelerates development but also enhances the adaptability and performance of intelligent and complex systems, making it an increasingly relevant approach to addressing the complexities of modern intelligent systems. Full article

Traditional systematic reviews, despite their high-quality evidence, are labor-intensive and error-prone, especially during the abstract screening phase. This paper investigates the application of machine learning-assisted systematic reviewing in the context of Learning Analytics (LA) in higher education. This study evaluates two approaches—ASReview, an active traditional machine learning tool, and GPT-4o, a large language model—to automate this process. By comparing key performance metrics such as sensitivity, specificity, accuracy, precision, and F1-score, we assess the effectiveness of these tools against traditional manual methods. Our findings demonstrate the potential of machine learning to enhance the efficiency and accuracy of systematic reviews in learning analytics. Full article

Background: Diabetes and mental health conditions frequently co-occur, with depression and anxiety affecting up to 20–30% of people with diabetes. These comorbidities worsen glycemic control, adherence, and quality of life, yet mental health is often neglected in diabetes care. Integrating mental health services into diabetes management is recommended by international organizations to improve patient outcomes. Objectives: To systematically review the evidence on integrated mental health interventions in diabetes care, compared to usual diabetes care, in improving patient outcomes (glycemic control, mental health, adherence, quality of life). Methods: We searched PubMed/MEDLINE, Embase, PsycINFO, and Scopus (2000 through July 2024) for studies of diabetes care integrating mental health support (e.g., collaborative care, co-location, stepped care, or digital interventions). Inclusion criteria were controlled trials or cohort studies involving individuals with type 1 or type 2 diabetes receiving an integrated mental health intervention, with outcomes on glycemic control and/or mental health. Two reviewers independently screened titles/abstracts and full texts, with disagreements resolved by consensus. Data on study design, population, intervention components, and outcomes were extracted. Risk of bias was assessed using Cochrane or appropriate tools. Results: Out of records identified, 64 studies met inclusion criteria (primarily randomized controlled trials). Integrated care models consistently improved depression and anxiety outcomes and diabetes-specific distress, and yielded modest but significant reductions in glycated hemoglobin (HbA1c) compared to usual care. Many interventions also enhanced treatment adherence and self-management behaviors. For example, collaborative care trials showed greater depression remission rates and small HbA1c improvements (~0.3–0.5% absolute reduction) relative to standard care. Co-located care in diabetes clinics was associated with reduced diabetes distress, depression scores, and HbA1c over 12 months. Digital health integrations (telepsychiatry, online cognitive-behavioral therapy) improved psychological outcomes and adherence, with some reporting slight improvements in glycemic control. Integrated approaches often increased uptake of mental health services (e.g., higher referral completion rates) and showed high patient satisfaction. A subset of studies reported fewer emergency visits and hospitalizations with integrated care, and one economic analysis found collaborative care cost-effective in primary care settings. Conclusions: Integrating mental health into diabetes care leads to better mental health outcomes and modest improvements in glycemic control, without adverse effects. Heterogeneity across studies is noted, but the overall evidence supports multidisciplinary, patient-centered care models to address the psychosocial needs of people with diabetes. Healthcare systems should prioritize implementing and scaling integrated care, accompanied by provider training and policy support, to improve outcomes and bridge the persistent treatment gap. Future research should focus on long-term effectiveness, cost-effectiveness, and strategies to reach diverse populations. Full article

Rett syndrome (RTT) presents with a wide range of symptoms spanning various clinical areas. Capturing symptom change as the disorder progresses is challenging. Wearable sensors offer a non-invasive and objective means of monitoring disease states in neurodevelopmental disorders. The goal of this study was to conduct a systematic literature review to critically appraise the literature on the use of wearable sensors in individuals with RTT. The PRISMA criteria were used to search four databases without time restriction and identified 226 records. After removing duplicates, the titles and abstracts of 184 records were screened, 147 were excluded, and 37 were assessed for eligibility. Ten (10) articles remained, and a further two were included after additional searching. In total, 12 articles were included in the final analysis. The sample size ranged from 7 to 47 subjects with an age range of 1 to 41 years. Different wearable biosensor devices were used across studies, with the Empatica E4 wearable device being most frequently used in 33% (4/12) of the studies. All the studies demonstrated a high methodological quality with a low risk of bias. Evidence from wearable sensors, combined with machine learning methods, enabled the prediction of different sleep patterns and clinical severity in RTT. Given the small sample size and the limitations of available data for training machine learning models, we highlight areas for consideration. The review emphasises the need to enhance research on the application of wearable sensors in epilepsy and gastrointestinal manifestations/morbidity in RTT. Increased electrodermal activity (EDA), % of maximum heart rate (HRmax%) and the heart rate to low-frequency power (HR/LF) ratio were identified as physiological measures potentially associated with disease states. Based on the evidence synthesis, the role of physiological parameters and their association with symptom management in RTT is discussed. Full article

The exponential growth of scientific literature demands scalable methods to evaluate large-language-model outputs beyond surface-level fluency. We present a two-phase framework that separates generation from evaluation: a retrieval-augmented generation system first produces candidate abstracts, which are then embedded into semantic co-occurrence graphs and assessed using seven robustness metrics from complex network theory. Two experiments were conducted. The first varied model, embedding and prompt configurations, achieved results showing clear differences in performance; the best family combined gemma-2b-it, a prompt inspired by chain-of-Thought reasoning, and all-mpnet-base-v2, achieving the highest graph-based robustness. The second experiment refined the temperature setting for this family, identifying $\tau =0.2$ as optimal, which stabilized results (sd $=0.12$) and improved robustness relative to retrieval baselines ($\Delta E_G=+0.08$, $\Delta \rho =+0.55$). While human evaluation was limited to a small set of abstracts, the results revealed a partial convergence between graph-based robustness and expert judgments of coherence and importance. Our approach contrasts with methods like GraphRAG and establishes a reproducible, model-agnostic pathway for the scalable quality control of LLM-generated scientific content. Full article

Background: Green tea (Camellia sinensis) has been consumed for centuries, and its beneficial effects on human health have been studied in recent decades. l-theanine, an active ingredient in green tea, has been used to improve cognition and mood. Although the effects of l-theanine on cognition have been investigated in clinical trials that have reported various results, these studies have not yet been critically evaluated in meta-analyses. Objectives: Our objective was to systematically evaluate the efficacy of l-theanine on cognitive functions compared to a placebo, in a meta-analysis based on randomized controlled trials (RCTs). Methods: PubMed, the Cochrane Central Register of Controlled Trials, Embase and Web of Science were searched for relevant studies until 31 July 2024 and registered in PROSPERO (registration number: CRD42024575122). Placebo-controlled clinical trials investigating the efficacy of l-theanine in healthy adults were included. Conference abstracts, study protocols and reports of non-RCTs were excluded. For risk of bias assessment, the Cochrane Risk of Bias Tool (version 2.0) was used. A random effects model was applied to conduct the meta-analysis. Mean differences (MD) with 95% confidence intervals (CIs) were calculated. Results: Based on the included five RCTs involving 148 healthy adults, l-theanine had a dose-dependent effect on cognitive function based on rapid visual information processing and recognition visual reaction time (MD: −15.20 ms; 95%-CI [−28.99; −1.41]). The effects of l-theanine were non-significant on reaction time to a simple stimulus (MD: −0.46 ms; 95% [CI: −15.65; 14.73]) and in the Stroop test (MD: −37.38 ms; 95%-CI [−86.39; 11.62]). Conclusions: The beneficial effects of l-theanine on cognitive performance could not be confirmed by all test methods. The contradictory results could be explained by the fact that l-theanine only affects certain cognitive domains, but also by the low number of trials and the heterogeneity of the test preparations. Further trials using standardized products with larger sample sizes are required for the accurate assessment of efficacy. Full article

This article introduces the ATHENA competency model, a systemic framework designed to conceptualize and support the development of creativity and complex skills in professional and educational contexts. Creativity, increasingly seen as essential across sectors, requires the coordination of cognitive, motivational, emotional, social, and sensorimotor resources. ATHENA conceptualizes competencies as emergent, agentic behaviors, not static possessions, arising from the coordination of five dimensions: cognition, conation, knowledge, emotion, and sensorimotion. These are subdivided into 60 facets, each described across four progressive mastery levels, enabling fine-grained diagnosis and developmental roadmaps. To operationalize this framework, ATHENA includes three modules: Skills, which models the requirements of professional tasks; Profile, which analyzes learner populations and contextual constraints; and LEARN, a repertory of pedagogical activities linked to ATHENA facets. The article illustrates the system through two case studies of creative job activities—graphic design and workshop facilitation—demonstrating how ATHENA aligns abstract competencies with practical training interventions. The model bridges theoretical research in psychology, creativity, and education with instructional design. Future work aims to refine its applicability, scalability, and cross-cultural relevance. Full article

Despite the popularity of electroencephalograms (EEGs) as tools for assessing brain health, they can sometimes be abstract and prone to noise, making them difficult to interpret. The following work aims to implement a Quaternion Convolutional Neural Network (QCNN) to detect abnormal EEGs obtained from a database that includes both people with excellent mental health and individuals with different types of mental illnesses. Unlike other approaches in which the QCNN is used exclusively for image processing, in the present work, a unique architecture with mainly quaternionic layers is proposed, specifically designed for the classification of time-varying signals. Using the database “The TUH EEG Abnormal Corpus”, the signals are preprocessed using the Wavelet Transform, a mathematical tool capable of performing simultaneous time and frequency analysis, configured with a level 4 decomposition value. Subsequently, the results are subjected to a partial spectrogram-type treatment to integrate the energy parameter into the analysis. They are then conditioned in each of the elements of the quaternion and processed by the QCNN, leveraging quaternion algebra to maintain the relationships between its elements, both in the input and in the convolutional product. In this way, it is possible to obtain significant percentages in the precision, recall, and accuracy metrics with values higher than 77%. Its performance, which uses 4 times less computational memory, allows the QCNN to be considered an alternative for classifying EEG signals. Finally, a comparison of the proposed model was made with other architectures commonly used in the literature, as well as with developments in other research and with a hybrid model whose performance places it at the highest classification standard, not to mention the ability of the QCNN to preserve multi-channel dependencies in EEG signals in a more natural way, achieving parameter efficiencies by leveraging quaternion algebra, reducing the computational cost compared to real-valued CNNs. Full article

Background: Nursing research has evolved through different historical stages, from the initial development of theoretical models to today’s challenges involving advanced practice and emerging technologies. Within this context, Middle-Range Nursing Theories (MRNTs) play a crucial role as a bridge between abstract conceptual frameworks and clinical practice. However, their recent production appears limited. Aims: To identify MRNTs published in the last five years, determine the main thematic fields addressed, and analyze current trends in their development. Methods: A scoping review was conducted in accordance with PRISMA-ScR guidelines. Databases searched included MEDLINE, CINAHL, PsycINFO, EMBASE, and Education Research Complete (August 2025). Eligible studies were published within the last five years in journals indexed in the Journal Citation Reports and explicitly proposed an MRNT. Exclusion criteria encompassed non-nursing theories, secondary applications of existing models, and purely methodological studies. Results: From 1230 initial records, 18 articles met the inclusion criteria. The Revista Brasileira de Enfermagem accounted for the highest number of publications. The identified MRNTs predominantly addressed clinical diagnoses and phenomena such as heart failure self-care, overweight, occupational stress, peripheral tissue perfusion, and social support networks. Most theories were derived from established nursing models (Orem, Roy, Levine, Neuman, Watson). Despite thematic diversity, few MRNTs had undergone methodological validation. Conclusions: Recent MRNT development remains limited and geographically concentrated, with Brazil emerging as a leading contributor. Strengthening methodological validation, clinical integration, and international dissemination is essential, as MRNTs continue to be pivotal tools for advancing nursing science, reinforcing disciplinary identity, and reducing the persistent gap between theory and practice. Full article

Background: Major advances in antiretroviral therapy (ART) have transformed HIV into a chronic condition, yet drug resistance, long-term toxicities, adherence challenges, and persistent viral reservoirs continue to drive innovation. Objectives: To map and synthesize recent developments in anti-HIV drugs and delivery platforms with a focus on (i) new molecules in clinical development and (ii) novel mechanisms of action, following a scoping review framework aligned with PRISMA-ScR. Sources: We interrogated PubMed, Embase.com, Web of Science, and Scopus (January 2020–September 2025) and screened abstracts from CROI, IAS/AIDS, IDWeek, and HIV Glasgow (2023–2025). Content: The evidence base underscores capsid inhibition (lenacapavir) for multidrug-resistant HIV and its expansion into prevention, long-acting intramuscular maintenance with cabotegravir/rilpivirine, maturation inhibitors (zabofiravir), and attachment inhibition with fostemsavir. Broadly neutralizing antibodies (bNAbs) can sustain ART-free suppression in selected individuals. Ultra-long-acting delivery systems are advancing toward translational evaluation. Summary: The pipeline is diversifying toward less frequent dosing, new targets, and combination strategies. Successful and ethical implementation will require resistance-informed selection, equitable access, and reimagined healthcare delivery models that accommodate long-acting technologies. Full article

The global transition toward sustainable energy has significantly accelerated the deployment of solar power systems. Yet, the inherent variability of solar energy continues to present considerable challenges in ensuring its stable and efficient integration into modern power grids. As the demand for clean and dependable energy sources intensifies, the integration of artificial intelligence (AI) with solar systems, particularly those coupled with energy storage, has emerged as a promising and increasingly vital solution. It explores the practical applications of machine learning (ML), deep learning (DL), fuzzy logic, and emerging generative AI models, focusing on their roles in areas such as solar irradiance forecasting, energy management, fault detection, and overall operational optimisation. Alongside these advancements, the review also addresses persistent challenges, including data limitations, difficulties in model generalization, and the integration of AI in real-time control scenarios. We included peer-reviewed journal articles published between 2015 and 2025 that apply AI methods to PV + ESS, with empirical evaluation. We excluded studies lacking evaluation against baselines or those focusing solely on PV or ESS in isolation. We searched IEEE Xplore, Scopus, Web of Science, and Google Scholar up to 1 July 2025. Two reviewers independently screened titles/abstracts and full texts; disagreements were resolved via discussion. Risk of bias was assessed with a custom tool evaluating validation method, dataset partitioning, baseline comparison, overfitting risk, and reporting clarity. Results were synthesized narratively by grouping AI techniques (forecasting, MPPT/control, dispatch, data augmentation). We screened 412 records and included 67 studies published between 2018 and 2025, following a documented PRISMA process. The review revealed that AI-driven techniques significantly enhance performance in solar + battery energy storage system (BESS) applications. In solar irradiance and PV output forecasting, deep learning models in particular, long short-term memory (LSTM) and hybrid convolutional neural network–LSTM (CNN–LSTM) architectures repeatedly outperform conventional statistical methods, obtaining significantly lower Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and higher R-squared. Smarter energy dispatch and market-based storage decisions are made possible by reinforcement learning and deep reinforcement learning frameworks, which increase economic returns and lower curtailment risks. Furthermore, hybrid metaheuristic–AI optimisation improves control tuning and system sizing with increased efficiency and convergence. In conclusion, AI enables transformative gains in forecasting, dispatch, and optimisation for solar-BESSs. Future efforts should focus on explainable, robust AI models, standardized benchmark datasets, and real-world pilot deployments to ensure scalability, reliability, and stakeholder trust. Full article

Background: Artificial intelligence (AI), the overarching field that includes machine learning (ML) and its subfield deep learning (DL), is rapidly transforming clinical research by enabling the analysis of high-dimensional data and automating the output of diagnostic and prognostic tests. As clinical trials become increasingly complex and costly, ML-based approaches (especially DL for image and signal data) offer promising solutions, although they require new approaches in clinical education. Objective: Explore current and emerging AI applications in oncology and cardiology, highlight real-world use cases, and discuss the challenges and future directions for responsible AI adoption. Methods: This narrative review summarizes various aspects of AI technology in clinical research, exploring its promise, use cases, and its limitations. The review was based on a literature search in PubMed covering publications from 2019 to 2025. Search terms included “artificial intelligence”, “machine learning”, “deep learning”, “oncology”, “cardiology”, “digital twin”. and “AI-ECG”. Preference was given to studies presenting validated or clinically applicable AI tools, while non-English articles, conference abstracts, and gray literature were excluded. Results: AI demonstrates significant potential in improving diagnostic accuracy, facilitating biomarker discovery, and detecting disease at an early stage. In clinical trials, AI improves patient stratification, site selection, and virtual simulations via digital twins. However, there are still challenges in harmonizing data, validating models, cross-disciplinary training, ensuring fairness, explainability, as well as the robustness of gold standards to which AI models are built. Conclusions: The integration of AI in clinical research can enhance efficiency, reduce costs, and facilitate clinical research as well as lead the way towards personalized medicine. Realizing this potential requires robust validation frameworks, transparent model interpretability, and collaborative efforts among clinicians, data scientists, and regulators. Interoperable data systems and cross-disciplinary education will be critical to enabling the integration of scalable, ethical, and trustworthy AI into healthcare. Full article

In recent years, machine learning (ML) has been increasingly used in many fields, including medicine. Magnetic resonance imaging (MRI) is a non-invasive and effective diagnostic technique; however, manual image analysis is time-consuming and prone to human variability. In response, ML models have been developed to support MRI analysis, particularly in segmentation and classification tasks. This work presents an updated narrative review of ML applications in brain MRI, with a focus on tumor classification and segmentation. A literature search was conducted in PubMed and Scopus databases and Mendeley Catalog (MC)—a publicly accessible bibliographic catalog linked to Elsevier’s Scopus indexing system—covering the period from January 2020 to April 2025. The included studies focused on patients with primary or secondary brain neoplasms and applied machine learning techniques to MRI data for classification or segmentation purposes. Only original research articles written in English and reporting model validation were considered. Studies using animal models, non-imaging data, lacking proper validation, or without accessible full texts (e.g., abstract-only records or publications unavailable through institutional access) were excluded. In total, 108 studies met all inclusion criteria and were analyzed qualitatively. In general, models based on convolutional neural networks (CNNs) were found to dominate current research due to their ability to extract spatial features directly from imaging data. Reported classification accuracies ranged from 95% to 99%, while Dice coefficients for segmentation tasks varied between 0.83 and 0.94. Hybrid architectures (e.g., CNN-SVM, CNN-LSTM) achieved strong results in both classification and segmentation tasks, with accuracies above 95% and Dice scores around 0.90. Transformer-based models, such as the Swin Transformer, reached the highest performance, up to 99.9%. Despite high reported accuracy, challenges remain regarding overfitting, generalization to real-world clinical data, and lack of standardized evaluation protocols. Transfer learning and data augmentation were frequently applied to mitigate limited data availability, while radiomics-based models introduced new avenues for personalized diagnostics. ML has demonstrated substantial potential in enhancing brain MRI analysis and supporting clinical decision-making. Nevertheless, further progress requires rigorous clinical validation, methodological standardization, and comparative benchmarking to bridge the gap between research settings and practical deployment. Full article

We build on the predictive processing framework to show that intelligent behaviour is adaptive control, driven by accurate prediction and uncertainty reduction in dynamic environments with limited information. We argue that adaptive control arises through a process of re-concretisation, where learned abstractions are grounded in new situations via embodiment. We use this as an explanation of why AI models often generalise at the cost of detail while biological systems manage to tailor their predictions towards specific environments over time. On this basis, we utilise the notion of embodied prediction to provide a new distinction between biological intelligence and the performance illustrated by AI systems. Full article

As the complexity of Very-Large-Scale Integration (VLSI) circuits escalates, Design-for-Test (DFT) faces significant challenges. Traditional script-based automation flows are increasingly complex and present a high technical barrier for non-specialists. In order to overcome the above issue, this paper introduces DFTAgent, a novel framework that leverages Large Language Models to intelligently orchestrate a DFT toolchain. DFTAgent is evaluated on the ISCAS’85, ISCAS’89, and ITC’99 benchmarks. The results demonstrate that DFTAgent successfully completes the complete ATPG task cycle, achieving fault coverage comparable to a manually scripted baseline while exhibiting significant advantages in flexibility and error handling. By abstracting complex DFT tools behind a natural language interface and a visual workflow, this approach promises to democratize access to advanced VLSI testing methodologies and accelerate design cycles. Full article