155199 research outputs found
Sort by
Diagnostic Accuracy of Machine Learning Algorithms in Electrocardiogram-Based Heart Failure Detection: A Systematic Review and Meta-Analysis
No full textBackgroundArtificial intelligence (AI) applied to the electrocardiogram (ECG) shows promise for detecting heart failure (HF), but reported performance is heterogeneous. A key ambiguity is the conflation of two distinct diagnostic targets: the structural abnormality of left ventricular systolic dysfunction (LVSD) and the clinical syndrome of HF.MethodsFollowing PRISMA-DTA guidelines, this systematic review and meta-analysis analyzed 40 unique, non-overlapping patient cohorts. Diagnostic accuracy was synthesized using a hierarchical bivariate model, addressing ejection fraction (EF) threshold heterogeneity via stratification and multi-threshold analysis. Prespecified bivariate meta-regressions examined covariates including external validation status, lead configuration, and model architecture. A secondary analysis evaluated HF classification models.ResultsThe primary analysis yielded a pooled sensitivity of 85.9% (95% CI 82.8–88.5%) and specificity of 80.9% (95% CI 75.8–85.1%), with a hierarchical summary receiver operating characteristic area under the curve (HSROC AUC) of 0.902 (95% CI 0.885–0.915). Performance varied significantly by target definition (LVSD vs. clinical HF) and EF threshold used. Meta-regression revealed that 12-lead ECGs (p=0.003) and convolutional neural network architectures (p=0.024) were associated with higher specificity. The secondary analysis (7 studies) yielded pooled sensitivity of 96.2% and specificity of 92.1%.ConclusionsAI-ECG demonstrates substantial but variable diagnostic performance that depends critically on target condition definition, EF thresholds, and methodological factors. Implementation must account for these dependencies and utilize precise, standardized endpoints.</div
Adapting the Short Digital Stress Scale (SDSS) into Turkish: validation and psychometric evaluation in adults
No full textNext-generation nutrition: Innovative and AI-tailored concentrated ingredients
No full textBackground: The growing demand for nutrient-dense and personalized foods has accelerated research into concentrated nutritional ingredients derived from diverse biological sources. Dairy streams, plant-based proteins, algae, fungi, seaweeds and edible insects provide abundant opportunities to extract high-value compounds through advanced separation and fermentation processes. Technologies such as microfiltration, ultrafiltration, nanofiltration, and precision fermentation have transformed conventional food processing into systems capable of producing bioactive, functional, and sustainable ingredients. However, optimizing these complex bioprocesses requires tools that can manage large datasets and predict multifactorial outcomes. In this context, artificial intelligence (AI) and machine learning (ML) are emerging as powerful enablers that can design, model, and control processing parameters to enhance yield, stability, and nutrient bioavailability. Scope and approach: This review synthesizes current advances in the development of concentrated nutritional ingredients from conventional (dairy and coproducts) and novel (plant, fungal, and insect) sources. It examines how AI and ML technologies can optimize the associated bioprocesses, including membrane fractionation, bioengineering, and precision fermentation, to improve efficiency, sustainability, and personalization. Key findings and conclusions: Concentrated nutrition represents a pivotal step toward achieving precision and sustainability in modern food systems. AI-driven bioprocess optimization enables data-informed control of nutrient extraction, functionality, and formulation, bridging the gap between biological potential and tailored human nutrition. Despite promising advances, challenges remain in model transparency, consumer acceptance, regulatory frameworks, and the scalability of AI-assisted production. Future progress will rely on the development of interoperable data systems, ethical frameworks, and cross-disciplinary collaboration to transform biological resources into intelligent nutrition systems for the next decade
Synthesis and in vitro/in silico evaluation of novel aroyl thiourea derivatives based on substituted phenethylamine: antibacterial, antioxidant, and anticancer activities
No full textThiourea derivatives possess a wide spectrum of biological activities and play a crucial role in drug design due to their anticancer, antimicrobial, antiviral, and enzyme inhibitory properties. In this study, seven novel substituted phenethylamine-based aroyl thiourea derivatives (23–29) with previously unreported structures were synthesized, and their biological activities were comprehensively evaluated. In the synthesis step, various substituted phenethylamines were reacted with 4-methoxyphenyl chloroformate in the presence of potassium thiocyanate, and the resulting compounds were characterized using FT-IR and NMR spectroscopic techniques. Antibacterial assays revealed that compounds 25, 26 , and 28 exhibited strong antibacterial activity against multidrug-resistant pathogenic strains, with MIC values ranging from 6.25 to 100 μM. Notably, compounds 25, 27, and 28 displayed the most potent efficacy with MIC values as low as 6.25 μM. Antioxidant analyses performed using the CUPRAC and DPPH methods demonstrated that electron-donating groups (compound 27) enhanced reducing capacity, whereas electron-withdrawing groups (compound 28) limited this effect. In anticancer evaluations, compounds 27 and 28 exhibited significant antiproliferative activity against A549 cells with IC50 values of 124.4 μM and 106.1 μM, respectively, while displaying minimal cytotoxicity toward normal HDF-1 cells. In silico analyses indicated that all synthesized compounds complied with Lipinski's rule of five and possessed favorable bioavailability. Molecular docking studies further confirmed these findings, revealing strong binding affinities ranging from −5.3 to −9.2 kcal/mol. Specifically, compound 26 showed the highest binding potential with a score of −9.2 kcal/mol against the target protein (PDB ID: 5AA4). These findings suggest that the synthesized aroyl thiourea derivatives serve as promising multifunctional biological agent candidates
Bibliometric Analysis of Medication Adherence in Transplantation
No full textBackground: Adherence to post-transplant immunosuppressive medications is one of the most critical factors in maintaining long-term graft survival. This study aimed to conduct a bibliometric analysis of medication adherence in transplantation. Methods: The data for this study was screened from the Scopus database. Bibliometric analysis and data visualization were conducted using VOSviewer, employing scientific mapping and performance analysis techniques. Results: A total of 1552 publications were analyzed, including 1263 articles and 289 reviews involving 8250 authors. These publications appeared in 619 different journals. The journals “Pediatric Transplantation” and “Clinical Transplantation” featured the most articles. The United States and the United Kingdom were the leading countries regarding the number of articles published. The most frequently cited author was De Geest (n = 44). Commonly used keywords included “adherence,” “medication adherence,” and “kidney transplantation.”. Conclusions: This study provides a comprehensive review of the literature on medication adherence in organ transplantation, identifying the most influential studies, leading journals, productive countries, institutions, and authors, and mapping research trends. The findings may guide healthcare professionals in the field. International collaborations with leading institutions and authors could help develop guidelines to improve adherence. The most cited studies have focused on the clinical and economic consequences of nonadherence. Future research is recommended to focus on examining the impact of health policies on medication adherence
Reduction in peripheral expression of the TMLHE gene in Turkish youth with autism spectrum disorder
No full textBackground: Trimethyllysine Hydroxylase, Epsilon (TMLHE) gene mutations have been clinically associated with an increased risk of autism spectrum disorder (ASD). This study aimed to evaluate the peripheral expression profile of the TMLHE gene and its association with ASD phenotype in a clinical sample of youth diagnosed with ASD. Methods: The study sample included 205 participants (ASD: n = 100; controls: n = 105, Mage = 9.25 years, SD = 3.74). The Childhood Autism Rating Scale and the Aberrant Behavior Checklist were administered to assess the severity of ASD and associated symptoms. Peripheral blood samples were collected from all participants, and TMLHE gene expression levels were analyzed using quantitative reverse transcription PCR (RT-qPCR). Results: TMLHE gene expression was significantly downregulated in the ASD group compared to controls (p < .001). Notably, significant correlations were identified between TMLHE expression levels and the CARS subscales for object use (p = .043) and listening response (p = .038). Conclusion: This study represents the first case-control investigation of peripheral TMLHE gene expression in ASD, revealing that TMLHE expression is reduced in children with ASD compared to typically developing peers. These findings contribute to a deeper understanding of the potential implications of TMLHE gene mutations in the etiology of ASD
Introducing AES and ECDSA to BLE Communication for Enhanced Security
No full textBluetooth Low Energy (BLE) has recently seen widespread adoption among Internet of Things (IoT) applications, sensor networks, personal area networks, and automation systems, thanks to its “lightweight" nature and low energy consumption. However, it lacks the most advanced security measures, which raises security concerns for use in critical scenarios. This work aims to integrate higher security standards into the BLE protocol, namely by ensuring confidentiality via Advanced Encryption Standard (AES) and by enabling integrity and non-repudiation through the Elliptic Curve Digital Signature Algorithm (ECDSA). The proposed framework was validated and tested over a testbed containing ESP32s with a BLE module, a camera, and an LCD screen. While the communication and computation overhead increases due to cryptographic redundancies, optimization methods were offered to minimize the impact. Hence, the desired security levels in confidentiality, integrity, and non-repudiation were successfully achieved within an acceptable performance margin for most IoT setups, making BLE more useful in critical scenarios
