Yayınlar
Koleksiyondaki öğeler
- MakaleAçık Erişim2025Scopus SJRQ2Bending and Buckling Analysis of Porous 2D Functionally Graded Beams with Exponential Material Property Variation
Adiyaman, Gökhan; Turan, Muhittin
The bending and buckling analysis of porous two-directional (2D) functionally graded (FG) beams was conducted using a higher-order shear deformation theory. The introduction of exponential functions to depict changes in material properties is a novel approach in the static analysis of 2D FG beams. Three distinct porosity distribution functions were taken into account. The governing equations were formulated through the application of Lagrange’s principle. During the numerical analysis, a finite element comprising two nodes and eight degrees of freedom was utilized. This choice facilitated accurate and efficient solutions, even for shorter beams, without the need for a shear correction factor. Notably, the obtained shear stresses aligned with actual values, registering as zero at both the top and bottom of the beam. The obtained results of the study were validated against findings reported in the literature. A parametric study was carried out to investigate the effects of porosity, porosity distributions, gradation parameters, slenderness, and boundary conditions on the non-dimensional deflections, stresses, critical buckling loads, and buckling mode shapes. It was found that both porosity and the distribution of porosity have noticeable effects on the static analysis of the beams.
AtıfScopus18OpenAlex13
İncele - MakaleYayıncıda açık erişim2026Scopus SJRQ1Optimization of extraction conditions of Micromeria fruticosa and use of optimized extract in beef patties production
Anlar Daldal, Pınar; Cinar Topcu, Kübra; Çakır, Özlem
This study aimed to optimize the conditions for obtaining an ethanol extract from Micromeria fruticosa (M. fruticosa) and to determine the antioxidant activity and phenolic compound content of the resulting extract. Subsequently, the use of the optimized extract in beef patties was evaluated, and the effects of the cooking process on the physicochemical properties of the product were examined. In this context, the highest efficiency was obtained in terms of total phenolic contents (TPC), as well as DPPH (2,2-diphenyl-1-picrylhydrazyl) and ferric reducing antioxidant power (FRAP) values, during the ethanol extraction of the plant. The optimum extraction conditions were determined to be 12 g of sample, 20 min, and 75% ultrasonic power. Under these conditions, the TPC of the extract was found to be 37.128 mg GAE/g dry weight, the DPPH value was 42.865, and the FRAP value was 254.776 mg TE/g dry weight. In the second phase of the study, the optimized extract was used at different concentrations (0%, 0.5%, 1.0%, and 1.5%) to produce beef patties, which were then analyzed for pH, moisture content, TBARS (thiobarbituric acid-reactive substances), and cooking yield. The use of the extract did not cause a statistically significant difference in the moisture content and TBARS values of the samples (p > 0.05). However, it was determined that the pH values of the samples containing 0% and 0.5% extract were lower than those containing 1.0% and 1.5% extract (p < 0.01). In addition, it was observed that the use of the extract had a positive effect on cooking yield, and that cooking yield increased with increasing extract concentration. As a result of the study, it was concluded that M. fruticosa holds promising potential as a natural additive in meat and meat products, offering a functional ingredient that can contribute to the development of value-added, health-oriented food products.
- MakaleYalnız künye2025Scopus SJRQ2Influence of ultrasonic-assisted deep eutectic solvent extraction on bioactive compounds from germinated mung beans (Vigna radiata) and greenness assessment
Yücel, Tuğçenur; Kamiloğlu, Aybike; Kutlu, Naciye
In this study, the ultrasonic-assisted method was applied to extract phenolic, flavonoid, and antioxidant compounds from germinated mung beans. In addition, deep eutectic solvents (DESs) were chosen as solvents to support the green concept and to increase their applicability. Choline chloride is a quaternary ammonium salt and was used as the hydrogen bond acceptor (HBA), and compounds from the alcohol, organic acid, sugar, and amide groups were combined as hydrogen bond donors (HBD) to obtain 8 different DESs. Higher yields were obtained with ultrasonic-assisted DESs compared to the traditional method (maceration) when total phenolic compound and antioxidant activities were evaluated. While the most effective solvent for total phenolic compound yield was choline chloride:fructose, the highest antioxidant activity values were obtained with solvents using organic acids as HBD. Moreover, Blue Applicability Grade Index (BAGI) was used to assess the greenness of the applied ultrasonic-assisted spectrophotometric extraction and DESs. As a result, using green technology and green solvents was influential in the extraction of bioactive compounds, and their applicability was especially effective for mung beans.
AtıfScopus8WoS7OpenAlex8
İncele - MakaleAçık Erişim2026Scopus SJRQ3Testing the Relationship between Investor Risk Appetite and Country Risk: The Case of Turkey
Kaya, Abdulkadir; Gulhan, Unal; Gungor, Bener
Summary: Country risk is an important factor affecting the risk appetite of investors who want to create an optimal portfolio by minimizing their risks. Investor risk appetite can affect country risk as a factor contributing to the development of financial markets and the country’s economy. The aim of this study is to determine the relationship between investor risk appetite and country risk. Hatemi-J cointegration and Hatemi-J asymmetric causality tests were used to determine the relationship between CDS premium, which is frequently preferred to represent country risk, and domestic and foreign investor risk appetite variables. As a result of the analysis, it has been determined that CDS premium and domestic and foreign investor risk appetite variables are cointegrated, a positive causality from increases in both local and foreign investor risk appetite to CDS premium and a negative or positive causality from decreases in CDS variables to both local and foreign investor risk appetite.
AtıfWoS1OpenAlex2İncele - BildiriAçık Erişim2025Scopus SJRQ4Lightweight Hybrid CNN Model for Face Presentation Attack Detection
Turhal, Uğur; Yilmaz, Asuman Günay; Nabiyev, Vasif
Today, face recognition systems are widely used in many areas that require biometric-based verification, especially because they are contactless and require low user cooperation. Despite their ease of implementation, these systems are vulnerable to attacks. Especially the increasing use of social media, makes it easier to spoof face recognition systems. Therefore, it is very important to develop new methods for Face Presentation Attack (FPA) detection. In this study, FPA detection performances of the lightweight Convolutional Neural Network (CNN) models -MobileNetV3-Small, ShuffleNet, SqueezeNet- and lightweight hybrid CNN models designed with combinations of these networks were investigated. In the experiments, the change in FPA detection performance was examined by giving the cropped face regions and the whole images as input to the networks. Using the whole image instead of the cropped face region significantly improved the FPA detection performance of the CNN models. These results showed that non-facial areas carry important information in FPA detection. Also, the FPA detection performance of the proposed lightweight hybrid CNN models were better than the single use of the networks for both inputs. 0.17% HTER performance was achieved with the hybrid CNN model built from the combination of ShuffleNet + SqueezeNet on the Replay-Attack dataset.
- Derleme MakaleAçık Erişim2026Scopus SJRQ1Artificial Intelligence Supported Online Assessment and Evaluation in Higher Education: A Systematic Review With SWOT Analysis
Üstün, Ayşin Gaye; Kayalı, Bünyami; Yavuz, Mehmet
Artificial intelligence-supported online assessment and evaluation (AI-SOAE) systems have the potential to transform assessment and evaluation processes in higher education; however, there is a need for a comprehensive evaluation of the pedagogical, technical, and ethical dimensions of these systems. In this regard, the study systematically examined the characteristics of AI-SOAE systems within the framework of a SWOT analysis. A systematic literature review method was used in the research. Ninety-eight studies published between 2018 and 2025 in the Web of Science (WoS), Scopus, and Education Resources Information Centre (ERIC) databases were analysed in accordance with PRISMA guidelines. The findings show that the strengths of AI-SOAE systems include personalised feedback, advanced assessment processes, increased efficiency, and data-driven decision support mechanisms. In contrast, technical infrastructure limitations, algorithmic biases, data management issues, and negative user experiences were identified as the weaknesses of the systems. The analysis also reveals that AI-SOAE systems offer significant opportunities in terms of pedagogical transformation, digitalisation, and institutional competitiveness; however, they face threats such as data security, ethical concerns, risks to academic integrity, and limited empirical evidence. Consequently, AI-SOAE systems offer context- and application-dependent potential for supporting assessment and evaluation practices in higher education. Effectively realising this potential depends on strengthening technical infrastructure, managing algorithmic risks, and developing transparent, ethical, and institutional policy frameworks. By synthesising the literature in the field of AI-SOAE in a structured manner, this study provides a comprehensive assessment for researchers, educators, and policymakers. © The Author(s) 2026. This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 License (https://creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).
AtıfScopus1WoS1OpenAlex3İncele - MakaleAçık Erişim2026Scopus SJRQ2Dual-Band Electromagnetic Shielding of Concrete Block Walls Using Concentric Double Square-Ring Frequency-Selective Surfaces near the n78 and n79 5G Frequency Regions
Cakir, Mehmet
Featured Application: This study demonstrates a concrete-integrated dual-band frequency-selective shielding approach for selectively attenuating the main sub-6 GHz 5G n78 and n79 frequency regions in building walls without increasing wall thickness or altering conventional concrete block geometry. This paper presents a dual-band shielding approach in which two concentric square rings, fabricated on a single 0.2 mm copper sheet, are embedded in the mortar joint of prefabricated concrete blocks. The outer ring targets n78 (design center 3.5 GHz) and the inner ring targets n79 (design center 4.7 GHz). Both rings are dimensioned from the effective wavelength inside the concrete medium (εr = 5.0): the outer ring has an outer side of 9.93 mm and the inner ring an outer side of 7.43 mm, with a 180° gap offset between the two elements to suppress inter-ring coupling. CST Microwave Studio simulations predicted two shielding peaks of 37.2 dB at 3.48 GHz and 35.8 dB at 4.72 GHz, with −10 dB bandwidths of 580 MHz and 490 MHz, respectively. Computer numerical control (CNC)-milled copper panels cast into ordinary Portland cement (OPC) concrete blocks achieved measured peak SE of 34.9 dB (n78) and 33.4 dB (n79), with bandwidths of 510 MHz and 420 MHz. The 2.3–2.4 dB shortfall relative to simulation is attributed to coupling-gap drift during milling and spacer-induced surface voids. A permittivity sweep from εr = 4.5 to 6.0 showed that peak SE remains above 30 dB throughout.
- MakaleYalnız künye2026Scopus SJRQ1Structural and physicochemical characterisation of branched dextrans produced by an active α-(1→2) branching sucrase from Apilactobacillus kunkeei PDER37
İspirli, Hümeyra; Dertli, Enes
Recently, branching sucrases encoded in the genomes of certain Lactic Acid Bacteria (LAB) strains have become novel enzymes to obtain branched α-glucans. In this study an active α-(1 → 2) branching sucrase from Apilactobacillus kunkeei PDER37 was expressed, characterised and distinct branched dextrans was obtained with reactions under different sucrose: dextran ratio. Structural characterisation by 1H and 13C NMR analysis demonstrated the branching of the dextran with (1 → 2)-linked α- d -glucose units with no alteration in the final structure depending on sucrose: dextran ratio (D0) but this ratio was effective for the determination of the molecular weights of the branched dextrans (D1, D2 and D3). FTIR analysis further supported the dextran structures and suggested the higher accumulation of the α-Glc units in the branched dextrans. Thermal characterisation of the branched dextrans obtained by TGA and DSC analysis suggested the increased hygroscopicity of the branching units. Both SEM and AFM analysis demonstrated more porous chain like structures in the branched dextrans. This study provides valuable information on the role of active α-(1 → 2) branching sucrase (BS37) for the production of branched dextrans with potential increased physicochemical status applicable for food and other industries.
- MakaleAçık Erişim2026Scopus SJRQ3DFT-based structural and energetic characterisation of nitrogen-rich tetrazine/triazine derivatives with Kamlet-Jacobs performance predictions
Ekincioğlu, Yavuz; Kepceoğlu, Abdullah; Çakir, Zehra; Şentürk, Şükrü
Nitrogen-rich heterocycles are attractive candidates for next-generation energetic materials because they can deliver high heats of formation while producing relatively benign, N2-rich decomposition products. In this work, three nitrogen-rich frameworks 3,6-bis(tetrazol-5-yl)−1,2,4,5-tetrazine (H2BTTz), 2,4,6-tris(tetrazol-5-yl)−1,3,5-triazine (H3TTT), and 2,3,5,6-tetra(tetrazol-5-yl)pyrazine (H4TTP) were investigated using a combined conformational sampling and density functional theory (DFT) approach. Initial conformational searches were carried out with the MMFF method, and the lowest-energy structures were optimised at the DFT/B3LYP/6-311++G(d,p) level a widely used compromise between accuracy and computational cost for nitrogen-rich energetic molecules. Electronic structure analyses included frontier molecular orbitals, Mulliken and natural population charges, and molecular electrostatic potential maps. Energetic performance was evaluated through heats of formation, detonation velocity and pressure, and impact sensitivity estimates. Topological descriptors such as ELF, LOL, and NCI were analyzed using Multiwfn program. The calculated HOMO–LUMO gaps (3.54-4.90 eV) indicate electronically stable molecular frameworks. Gas-phase formation enthalpies were converted to condensed-phase values using a sublimation correction, and detonation properties were estimated via the Kamlet–Jacobs equations. The predicted performance follows the order H4TTP > H2BTTz > H3TTT. Although their detonation performance is lower than that of FOX-7 and CL-20, these nitrogen-rich systems remain promising candidates for reduced-sensitivity energetic materials and merit further experimental investigation.
AtıfScopus1WoS1OpenAlex1İncele - MakaleYalnız künye2026Scopus SJRQ1Upcycling marble waste into construction materials and heavy metal adsorbents via hot-pressed geopolymerization technique
Öksüzer, Nurullah; Uslu, Emin; Gökçe, H. Süleyman
This study presents a sustainable approach to utilizing marble dust waste by producing bifunctional metakaolin-based geopolymers via a synchronized hot-pressing technique. In this study, the environmental footprint was minimized by designing mixtures in which marble waste constituted 50% of the total volume. Three stoichiometric series were investigated by varying the activator modulus (SiO2/Na2O = 3 and 4) and the Si/Al ratio (2 and 3) under two curing temperatures (150 and 300 °C) and three compaction stresses (5, 25, and 50 MPa). Thus, a high-volume industrial byproduct was transformed into a functional component. In this context, it investigated the transformation of industrial products into next-generation materials by using them to produce cementless composite geopolymers for civil engineering and to adsorb heavy metal (Cu2+). As a result of the research, an early compressive strength of 58 MPa was obtained through heat treatment and pressing applied at 300 °C and 50 MPa. This optimum strength was achieved in the SiO2/Na2O = 3, Si/Al = 3 series, whereas excess silica (SiO2/Na2O = 4) and excess aluminum (Si/Al = 2) limited the strength to 16.8 and 35.2 MPa, respectively. Microstructural analyses (SEM/EDS and FT-IR) revealed that marble dust was used as a filler material in the microstructure to produce an N-A-S-H-containing matrix. The produced materials were then ground after their service life and used to purify water containing heavy metals. More than 99.75% of Cu2+ was removed through ion exchange between copper ions used as heavy metals. The residual Cu2+ concentration fell below 0.25 mg/L from an initial concentration of 100 mg/L, and the released Na+-to-adsorbed Cu2+ molar ratio of approximately 1.8–2.3 confirmed an ion-exchange-dominated uptake mechanism. This structural and environmental impact demonstrates that marble waste, which is otherwise discarded as waste, is considered inert, offering an alternative model for the circular economy and environmental applications.
- Derleme MakaleYalnız künye2025Scopus SJRQ3Artificial Intelligence and Environmental Impact: Moving Beyond Humanizing Vocabulary and Anthropocentrism
Karakaş, Ümit; Özdemir, Vural
Artificial intelligence (AI) and its applications in digital health, bioengineering, and society have significant material impacts on the environment owing to AI’s vast energy demands and energy consumption, carbon footprints, and water usage to cool data centers and generate electricity to power the data centers. Yet, the environmental footprints of AI remain underappreciated and inadequately acknowledged. This is significant, particularly in this era of climate emergency and ongoing threats to planetary energy and water supplies. The vocabulary attached to AI often aims to mimic positive human capacities such as “warmness” and “care.” However, these attempts to humanize AI and digital technology come with an anthropocentric gaze and blind spots that bracket out the environmental impacts and footprints of AI and privilege humans and technology over nonhuman animals and planetary ecological limits. In medicine, the environmental impacts of large language models range from water consumption and carbon emission to rare mineral usage. This commentary and innovation analysis question and queer the popular imagination of AI and digital technology as things that only exist in the immaterial world of cyberspace. In the course of research on AI in planetary health, we must be cognizant of its materiality, ecological impacts, and massive energy and water demands. We argue that moving away from anthropocentric narratives and vocabulary in AI design and praxis would bode well to live within planetary ecological limits so that AI and emerging digital technologies best serve robust and responsible science and all life on the planet Earth. Copyright 2025, Mary Ann Liebert, Inc., publishers.
AtıfScopus10WoS7OpenAlex13
İncele - MakaleAçık Erişim2025Scopus SJRQ2Physicochemical properties, organic acid composition and free amino acid profiles in selected Cornus mas L. genotypes from the Çoruh Valley
ERDOĞAN, Ümmügülsüm; ERDOĞAN, Yaşar
Cornus mas L. (cornelian cherry) is a species widely used from traditional consumption to modern industrial applications due to its high nutritional content, richness in phenolic compounds, and functional food potential. This study aimed to comprehensively evaluate the physicochemical properties, organic acid composition, and free amino acid profiles of 15 cornelian cherry genotypes selected from the Çoruh Valley, one of Turkey's important biodiversity centers. Fruit samples were collected at full physiological maturity, and TSS, titratable acidity, pH, sugar composition, ascorbic acid, and tannin contents were analyzed using standard methods.Organic acids were determined by HPLC-UV, and free amino acids were determined by HPLC with OPA/FMOC derivatization. All measured traits differed significantly among genotypes (p < 0.05). Malic acid ranged from 38.12 to 251.34 mg/100 g, and total organic acids from 588.89 to 1038.15 mg/100 g. Free amino acid contents ranged from 71.34 to 93.14 mg/100 g, with the highest values in the G7 and G5 genotypes. This study demonstrates that the Çoruh Valley is an important hotspot of genetic diversity for cornelian cherry and contains genotype groups suitable for different uses. Overall, the findings provide a biochemical basis for selecting genotypes for breeding, nutritional applications, and industrial processing.
- MakaleYalnız künye2026Scopus SJRQ1AI-enhanced disaster education in middle schools: A mixed study from social studies classes
Yetişensoy, Okan
This research aims to examine the educational potential of AI-assisted chatbots, the popular AI technology of recent years, in middle school disaster education. In this regard, six chatbot characters in human form, each with fictional stories, were developed for various disasters taught within the scope of 5th-grade Social Studies in Türkiye. While the practitioner teacher instructed on disaster-related curricular outcomes in the experimental group through activities that included the use of the relevant chatbots, the same curricular outcomes in the control group were taught using conventional pedagogical methods. The experimental results showed that the disaster knowledge of the experimental group students was significantly higher than that of the control group students. However, no significant difference was observed in students’ disaster awareness levels. In the semi-structured interviews, students expressed that chatbots promoted a more effective, interesting, and student-centered disaster education with their distinctive design and pedagogical features. However, they noted that chatbots had certain inadequacies in developing concrete experiences, as well as psychomotor skills necessary to address disaster risks. Findings from the study indicate that the relevant AI technology is effective in promoting disaster education processes, particularly in facilitating more effective learning. Nevertheless, achieving more complex competencies for mitigating risks requires a higher level of technological advancement. At this point, it is suggested that undertaking initiatives to combine chatbots with extended reality environments will be beneficial in making them a more effective tools in disaster education processes.
AtıfScopus2İncele - MakaleAçık Erişim2026Scopus SJRQ2Production of Gamma-Aminobutyric Acid by Levilactobacillus brevis AK19B: Strain Characterization and Process Optimization Using Box–Behnken Design
Memiş, Habibe; Kamiloğlu, Aybike
Levilactobacillus brevis AK19B strain demonstrated promising probiotic potential as a hydrophilic lactic acid bacterium with valuable phytase activity. It also exhibited strong antioxidant properties and maintained viability in bile salt-rich environments, suggesting robustness in gastrointestinal conditions. To maximize gamma-aminobutyric acid (GABA) production, a Box–Behnken design was applied to assess the effects of monosodium glutamate (MSG) concentration, NaCl concentration, pH, and temperature. The results showed that these factors significantly influenced GABA synthesis, with their interactions well-modeled by a quadratic response surface. Optimal conditions were identified as pH 5.136, MSG concentration of 297 mM, temperature at 26 °C, and NaCl concentration of 2.82%, under which the predicted GABA production reached 43.85 mg/mL. These findings highlight AK19B as a promising candidate for functional food applications and industrial GABA biosynthesis.
- MakaleAçık Erişim2026Scopus SJRQ2Bioinspired Sternal Implant Design for Generic Anatomical Reconstruction: An In Silico Framework for Material Selection and Biomechanical Validation
Kutbay, Işıl; Gerdan, Zeynep; Çolak, Murat; Tabak, Yasemin; Şensoy, Abdullah Tahir
The sternum protects the intrathoracic organs and contributes to chest wall mechanics, which makes reconstruction after tumor resection, trauma, or infection a demanding biomechanical problem. This study presents an in silico workflow for preselecting materials for sternal implants before physical prototyping. After a virtual resection, an anatomically conformal implant was designed and candidate biomaterials were screened in CES Selector using density, elastic modulus, fatigue strength, fracture toughness, toxicity, medical grade suitability, and MRI safety. A representative subset of the screened candidates was then compared by finite element modeling in terms of stress transfer and deformation. Seventeen candidates met the screening criteria. Ti-13Nb-13Zr showed an elastic modulus of about 80 GPa, and the titanium-based candidates showed deformation values of about 0.96 to 1.03 mm, whereas GF PEEK reached about 1.74 mm. The stress shielding index also showed that titanium-based materials remained on the implant-dominant side, while polymer-based materials shifted stress transfer toward bone. Taken together, the findings suggest that Ti-13Nb-13Zr offers the best overall balance for load-bearing sternal reconstruction, whereas PEEK-based systems may be more suitable within the present model for hybrid or adjunct designs. The proposed workflow can support early implant planning and guide future experimental and clinical studies.
- MakaleAçık Erişim2025Scopus SJRQ1From screens to minds: the mediating role of psychological well-being between digital reading and AI anxiety
Özbilen, Uğur; Banaz, Emrullah; Şahin, Tuğrul Gökmen
Introduction: This study investigates the mediating role of psychological well-being in the relationship between digital reading disposition and artificial intelligence (AI) anxiety among Turkish teachers. Addressing the growing concern of technology-related anxiety in education, the research explores how digital literacy and psychological resilience interact within a single structural model. Methods: A correlational research design was employed with data collected from 324 teachers. Participants completed the Digital Reading Disposition Scale, the Psychological Well-Being Scale, and the Artificial Intelligence Anxiety Scale. Bootstrapped structural equation modeling (SEM) was used to test the measurement model and the hypothesized mediation effect. Prior to SEM analysis, confirmatory factor analysis (CFA) was conducted to establish construct validity of the scales. Results: Findings revealed that digital reading disposition was positively associated with psychological well-being and negatively linked with AI anxiety. Psychological well-being was also negatively related to AI anxiety. Moreover, psychological well-being partially mediated the relationship between digital reading disposition and AI anxiety. Discussion: The results suggest that teachers with stronger digital reading dispositions experience higher psychological well-being, which in turn buffers against AI-related anxiety. These findings contribute novel theoretical insights to educational literature by integrating digital reading, psychological well-being, and AI anxiety into a single model. Practically, the study underscores the importance of fostering digital literacy and psychological resilience to mitigate technology-related anxiety in educational settings.
AtıfWoS1OpenAlex2İncele - MakaleAçık Erişim2025Scopus SJRQ2Mind diet adherence associated with lower Zonulin levels in older Parkinson's patients
Esgin, Özge; Tokçaer Bora, Hatice Ayşe; Gülbahar, Özlem; Demirkesen, Özlem Cemile; Ataoğlu, Nigar Esra Erkoç; Söke, Fatih; Özçelik, Ayşe Özfer
Aim: This study investigates the association between adherence to the MIND diet and serum zonulin levels in older adults with Parkinson's disease (PD). Methods: Data were collected through structured survey forms and blood samples. The survey assessed general information, 24-hour food recalls, anthropometric measurements, the MIND diet adherence and physical activity levels. Zonulin levels were analysed from blood samples. Results: The study included 57 older adults with PD in stages 1–3 on the Hoehn & Yahr scale (29 women, 28 men; mean age 70.2 ± 4.6 years). Zonulin levels were negatively correlated with dietary total antioxidant capacity (p < 0.05). In adjusted analyses, higher adherence to the MIND diet was associated with lower zonulin levels (β = −0.25). Conclusions: These findings suggest a novel link between adherence to the MIND diet and reduced zonulin levels. MIND diet could have potential benefits for neurological function and intestinal barrier integrity.
- MakaleAçık Erişim2026Scopus SJRQ2Deep Learning-Based Classification of Common Lung Sounds via Auto-Detected Respiratory Cycles
Engin, Mustafa Alptekin; Uzun Arslan, Rukiye; Senyer Yapici, İrem; Aras, Selim; Gangal, Ali
Chronic respiratory diseases, the third leading cause of mortality on a global scale, can be diagnosed at an early stage through non-invasive auscultation. However, effective manual differentiation of lung sounds (LSs) requires not only sharp auditory skills but also significant clinical experience. With technological advancements, artificial intelligence (AI) has demonstrated the capability to distinguish LSs with accuracy comparable to or surpassing that of human experts. This study broadly compares the methods used in AI-based LSs classification. Firstly, respiratory cycles—consisting of inhalation and exhalation parts in LSs of different lengths depending on individual variability, obtained and labelled under expert guidance—were automatically detected using a series of signal processing procedures and a database was obtained in this way. This database of common LSs was then classified using various time-frequency representations such as spectrograms, scalograms, Mel-spectrograms and gammatonegrams for comparison. The utilisation of proven, convolutional neural network (CNN)-based pre-trained models through the application of transfer learning facilitated the comparison, thereby enabling the acquisition of the features to be employed in the classification process. The performances of CNN, CNN and Long Short-Term Memory (LSTM) hybrid architecture and support vector machine methods were compared in the classification process. When the spectral structure of gammatonegrams, which capture the spectral structure of signals in the low-frequency range with high fidelity and their noise-resistant structures, is combined with a CNN architecture, the best classification accuracy of 97.3% ± 1.9 is obtained.
AtıfScopus2WoS2OpenAlex2İncele - MakaleYalnız künye2026Scopus SJRQ4Evaluating Large Language Models in Turkish Short Answer Scoring: Validity, Reliability, and Fairness Perspectives
Kara, Abdulkadir; Yıldırım, Serkan
This study examines the performance of large language models (LLMs) in Turkish short-answer assessments within the measurement and evaluation theory framework. The GPT, Gemini, Gemma, and LLaMA models were evaluated under zero-shot and one-shot conditions with rubric support. The results show that LLMs have high self-consistency, but decision reliability can vary depending on prompt format and example sensitivity. Formulating rubrics with clear and concrete performance indicators increases model-human alignment and assessment fairness. Furthermore, error analyses revealed that while models often exhibit systematic low-scoring tendencies, they also display an ‘inverted-U’ error pattern, showing higher reliability at score extremes but struggling significantly with evaluating partial knowledge (intermediate scores). The results indicate that LLMs can support teachers in formative assessment when properly structured rubrics are used, but ethical oversight and pedagogical responsibility remain indispensable in final decisions.
- MakaleAçık Erişim2026Scopus SJRQ1A comparative study of metaheuristic algorithms for thermoeconomic optimization of an Organic Rankine Cycle
Tayfur, Bilal; Turhal, Uğur; Tozlu, Alperen
The efficient recovery of waste heat from biogas power plants via Organic Rankine Cycles (ORC) is essential for sustainable energy but requires satisfying conflicting thermodynamic and economic objectives. This study addresses the need for robust optimization strategies by evaluating nine recent metaheuristic algorithms to identify the most effective methods for maximizing net power while minimizing system costs. A multi-objective thermoeconomic model utilizing the specific exergy costing method was developed for a toluene-based ORC and optimized using algorithms including Dung Beetle, Snake, and NSGA-II. The optimization reduced the unit electricity cost by 9.1% (from $0.088/kWh to $0.080/kWh), with the Dung Beetle algorithm exhibiting superior convergence (Hypervolume 25.90) and the Snake algorithm demonstrating the best solution uniformity (Spread 1.54) with computational speed (0.084 s). These findings underscore that selecting algorithms based on specific exploration-exploitation characteristics is critical for achieving cost-effective and thermodynamically efficient power generation designs.
AtıfOpenAlex2İncele
