Scopus İndeksli Yayınlar Koleksiyonu
Permanent URI for this collectionhttps://hdl.handle.net/20.500.12573/395
Browse
317 results
Search Results
Article Non-Contact Acoustic Screening for Sleep Apnea: A Subject-Aware Deep Learning Approach(Springer Science and Business Media Deutschland GmbH, 2026-02-11) Aygün Çakıroğlu, M.; Kizilkaya Aydoǧan, E.; Bolatturk, Ö.F.; Aydoğan, S.; Ismailoǧullari, S.; Delice, Y.Purpose: To explore the feasibility of using camera-derived, non-contact audio synchronized with PSG for clinically relevant sleep-apnea classification, and to benchmark compact deep models under a subject-aware design using a previously unstudied, real-world dataset. Methods: Thirty-two adults underwent simultaneous polysomnography (PSG) and camera-based non-contact audio recording. The synchronized audio segments were used to train and compare three compact deep-learning architectures (convolutional, attention-augmented, and transformer-based) under a subject-aware evaluation design that prevented identity leakage. Model performance and calibration were assessed at both segment and subject levels using standard statistical tests. Results: Subject-level evaluation was based on a very small, imbalanced test set of six subjects (one positive). Within this limited yet previously unstudied local dataset, the CNN_trans model achieved an apparent perfect ranking performance (AUC = 1.00; 95% CI 0.00–1.00), though this likely reflects the small, imbalanced test cohort, with recall = 1.00 and precision = 0.55. The wide confidence interval reflects substantial statistical uncertainty, and DeLong comparisons showed no significant AUC difference between CNN_trans and CNN_att (ΔAUC = − 0.042; p = 0.43). Conclusion: PSG-synchronized, non-contact audio supports accurate and well-calibrated sleep-apnea classification with compact deep models. This subject-aware evaluation suggests that contactless acoustic monitoring may have potential clinical relevance, motivating larger, multi-site validation. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2026.Article Citation - WoS: 1Two-Local Modifications of Sachdev-Ye Model With Quantum Chaos(American Physical Society, 2026-01-27) Hanada, M.; Van Leuven, S.; Oktay, O.; Tezuka, M.The Sachdev-Ye-Kitaev (SYK) model may provide us with a good starting point for the experimental study of quantum chaos and holography in the laboratory. Still, the four-local interaction of fermions makes quantum simulation challenging, and it would be good to search for simpler models that keep the essence. In this paper, we argue that the four-local interaction may not be important by introducing a few models that have two-local interactions. The first model is a generalization of the spin-SYK model, which is obtained by replacing the spin variables with SU(d) generators. Simulations of this class of models might be straightforward on qudit-based quantum devices. We study the case of d=3,4,5,6 numerically and observe quantum chaos already for two-local interactions in a wide energy range. We also introduce modifications of spin-SYK and SYK models that have similar structures as the SU(d) model (e.g., H=∑p,qJpqχpχp+1χqχq+1 instead of the original SYK Hamiltonian H=∑p,q,r,sJpqrsχpχqχrχs), which shows strongly chaotic features although the interaction is essentially two-local. These models may be a good starting point for the quantum simulation of the original SYK model. ©2026 American Physical Society.Article Densification-Induced Chemical Reorganization and Mechanical Enhancement in Amorphous Si2BC3N(Elsevier, 2026-02) Durandurdu, MuratThe atomistic mechanisms that govern the mechanical performance of amorphous silicon-boron carbonitride (SiBCN) ceramics remain insufficiently understood, particularly regarding the role of density. Here, we employ ab initio molecular dynamics simulations to elucidate the structural evolution and mechanical response of low-density (LDA, 2.20 g/cm3) and high-density (HDA, 2.53 g/cm3) amorphous Si2BC3N prepared via melt-quench. The HDA phase exhibits markedly higher atomic packing and network connectivity, accompanied by a nontrivial chemical reorganization. Densification significantly enhances heteronuclear bonding-especially Si-C coordination-while suppressing C-C and Si-Si homopolar bonds. These changes yield substantial mechanical strengthening: the HDA phase exhibits a 48% increase in bulk modulus (130 GPa vs. 88 GPa), along with elevated Young's (266 GPa) and shear (112 GPa) moduli. Our findings reveal a clear density-structure-property relationship in amorphous SiBCN, demonstrating that densification suppresses weak self-bonded motifs and promotes a robust, interconnected atomic network. This insight provides a pathway for designing high-performance amorphous SiBCN ceramics for extreme-environment applications.Article Supervised Learning-Driven Dead Band Control of Occupant Thermostats for Energy-Efficient Residential HVAC(Elsevier, 2026-03) Savasci, Alper; Ceylan, Oguzhan; Paudyal, SumitHeating, ventilation, and air conditioning (HVAC) systems play a crucial role in demand-side management (DSM) by shaping residential electricity consumption and enabling flexible, grid-responsive operation. Thermostats in HVAC systems regulate indoor temperature as part of a closed-loop control framework, typically incorporating a fixed temperature dead band-a range around the setpoint where no action is taken-to reduce energy use and prevent frequent cycling of the HVAC system. Although essential for efficiency and equipment longevity, fixed dead bands limit adaptability, as dynamically adjusting them under varying environmental conditions remains challenging for occupants. To address this limitation, we propose a machine learning (ML)-based dead band tuning framework that optimally adjusts thermostat settings in real time. The method integrates conventional optimization with data-driven modeling: a mixed-integer linear programming (MILP) model is first used to gen erate optimal dead band values under measured outdoor temperature records (diverse seasonal weather scenarios) which are then employed to train the ML-based predictor to learn a real-time discrete dead band decision policy that approximates the MILP-optimal hysteresis-aware decisions. Among the evaluated models, Random Forest demonstrates superior predictive performance, achieving a mean squared error (MSE) of 0.0399 and a coefficient of determination (R2) of 95.75 %.Article Deep-Learning Detection of Open-Apex Teeth on Panoramic Radiographs Using YOLO Models(Springer, 2025-12-23) Edik, Merve; Celebi, Fatma; Cukurluoglu, AykaganObjectivesThe use of deep learning in detecting teeth with open apices can prevent the need for additional radiographs for patients. The presented study aims to detect open-apex teeth using You Only Look Once (YOLO)-based deep learning models and compare these models.MethodsA total of 966 panoramic radiographs were included in the study. Open-apex teeth in panoramic radiographs were labeled. During the labeling process, they were divided into 6 classes in the maxilla and mandible, namely incisors, premolars, and molars. AI models YOLOv3, YOLOv4, and YOLOv5 were used. To evaluate the performance of the three detection models, both overall and separately for each class in the test dataset, precision, recall, average precision (mAP), and F1 score were calculated.ResultsYOLOv4 achieved the highest overall performance with a mean average precision (mAP) of 87.84% at IoU (Intersection over Union) 0.5 (mAP@0.5), followed by YOLOv5 with 85.6%, and YOLOv3 with 84.46%. Regarding recall, YOLOv4 also led with 90%, while both YOLOv3 and YOLOv5 reached 89%. Moreover, the F1 score was the highest for YOLOv4 (0.87), followed by YOLOv3 (0.86) and YOLOv5 (0.85).ConclusionsIn this study, YOLOv3, YOLOv4, and YOLOv5 were evaluated for the detection of open-apex teeth, and their mAP, recall, and F1 scores exceeded 84%. Deep learning-based systems can provide faster and more accurate results in the detection of open-apex teeth. This may help reduce the need for additional radiographs from patients and aid dentists by saving time.Article Citation - WoS: 2Citation - Scopus: 2Neuroinflammatory Human Brain Organoids Enable Comprehensive Drug Screening Studies: Fingolimod and Its Analogues in Focus(Bentham Science Publishing Ltd, 2025-10-08) Acar, Busra; Pepe, Nihan Aktas; Zivkovic, Aleksandra; Stark, Holger; Sen, AlaattinIntroduction The absence of physiologically relevant models for neuroinflammatory brain disorders, such as multiple sclerosis (MS), highlights the need for improved drug screening platforms. To bridge this gap, this study aimed to develop a human brain organoid (hBO) model incorporating essential neural cell types, including astrocytes, microglia, and oligodendrocytes.Methods hBOs were generated from H9 stem cells, and neuroinflammatory characteristics were elicited by lipopolysaccharide (LPS). The expression of specific neuronal and inflammatory markers was assessed through qRT-PCR, immunofluorescence staining (IFS), and ELISA.Results IFS of mature hBOs with anti-SOX2, anti-SATB2, anti-MAPT, anti-GFAP, anti-MBP, and anti-IBA1 antibodies and images collected with the confocal microscope confirmed the differentiation of H9 cells into cortical neurons, astrocytes, microglia, and oligodendrocyte cell types. Elevated GFAP, IBA1, NF-kappa B, and IL-6 levels, along with reduced CNPase expression with LPS treatment, were considered reflective of MS-like pathology and were used to test fingolimod and its derivatives. Fingolimod and all its derivatives, specifically ST-1505, decreased MAPT (2.1-fold in ELISA, 1.7-fold in IFS), GFAP (1.8-fold in IFS), TNF alpha (5.4-fold in qRT-PCR), and FABP (1.5-fold in ELISA) levels, and increased IL-10 (11-fold in qRT-PCR) and MBP (2.9-fold in IFS) levels.Discussion The present data collectively showed LPS to evoke neuroinflammation in the hBO model, while fingolimod and its derivatives, particularly ST-1505, exhibited significant anti-inflammatory and neuroprotective properties by counteracting these evoked changes in the hBO model.Conclusion The findings supported the applicability of brain organoids as a model system for drug screening studies for neuroinflammatory brain diseases.Article Crashworthiness Evaluation of 3D-Printed Hybrid-Design Multi-Cell Energy Absorbers Under Lateral Compression for Unmanned Aerial Vehicles(Springer Heidelberg, 2025-11-25) Atahan, M. Gokhan; Zeybek, Halil; Ozturk, SezginEnergy absorbers can be strategically integrated into critical areas of unmanned aerial vehicles to protect their structural integrity and electronic components in the event of an accident. In this study, hybrid-design multi-cell energy absorber configurations were proposed, and their crashworthiness performance and collapse mechanisms were comparatively analyzed. Hybrid energy absorbers were designed considering circular, square, hexagonal, and re-entrant unit cell geometries. The energy absorber configurations were produced via additive manufacturing. Compared to the single-cell circular energy absorber, the hybrid-design multi-cell approach resulted in a higher peak crushing force value, while offering considerable enhancements in other crashworthiness parameters. Configuration 3 is recommended for use in energy absorber applications in unmanned aerial vehicles due to its superior crashworthiness performance. Moreover, in hybrid-design multi-cell energy absorbers, the selection of layer geometries significantly influences deformation capability. Compared to the single-cell circular configuration (Configuration 1), Configuration 3 demonstrated superior crashworthiness performance by increasing the MCF, EA, and SEA values by 7.47, 4.47, and 1.41 times, respectively.Article Tuning Properties of Amorphous Boron Via Hydrogenation: An Ab Initio Study(Elsevier, 2026-01) Durandurdu, MuratAb initio simulations are employed to investigate the structural, mechanical, and electronic properties of hydrogenated amorphous boron (a-B:H) across a range of hydrogen concentrations (approximate to 6-21 at.%). The results indicate that pentagonal-like boron clusters constitute the primary structural motifs. The bonding environment consists of both B-H terminal bonds and B-H-B bridging bonds, with the fraction of bridging bonds ranging from 10 % to 16 %. Increasing the hydrogen content leads to a reduction in density and bulk modulus, accompanied by a systematic widening of the electronic band gap. These results demonstrate that hydrogen incorporation profoundly modifies the atomic structure, softens the network, and enhances the semiconducting character of a-B:H, highlighting the tunability of properties in boron-based amorphous materials.Article Citation - WoS: 1Citation - Scopus: 1A Comparative Study of Existing and Current On-Site Documentation of Anatolian Seljuk Kümbets(Elsevier Ltd, 2025-12) Güzelci, O.Z.; Türel, A.During the Anatolian Seljuk period (1077–1307), monumental tombs known as kümbets emerged as a distinct architectural typology in present-day Türkiye. 2D drawings of these structures, produced since the early 20th century, contain inconsistencies that necessitate verification and accurate documentation. This study digitally documents Anatolian Seljuk kümbets in 3D to generate updated 2D sections reflecting their current condition and compares these with previously published drawings. The methodology includes collecting available 2D sections, digitally documenting kümbets through field studies, generating new 2D sections from 3D models, and systematically comparing these datasets. Two image-based metrics are employed in the comparison: the Exact Pixel Match Ratio (EPMR), which evaluates pixel-level alignment, and the Structural Similarity Index Measure (SSIM), a standard indicator for visual similarity. The results provide a comparative framework for assessing previous drawings and present a verified, up-to-date dataset of kümbet sections for future research. © 2025 Elsevier B.V., All rights reserved.Article Citation - Scopus: 3Chaos in PID Controlled Nonlinear Systems(Korean Institute of Electrical Engineers, 2015) Ablay, G.Controlling nonlinear systems with linear feedback control methods can lead to chaotic behaviors. Order increase in system dynamics due to integral control and control parameter variations in PID controlled nonlinear systems are studied for possible chaos regions in the closed-loop system dynamics. The Lur’e form of the feedback systems are analyzed with Routh’s stability criterion and describing function analysis for chaos prediction. Several novel chaotic systems are generated from second-order nonlinear systems including the simplest continuous-time chaotic system. Analytical and numerical results are provided to verify the existence of the chaotic dynamics. © 2021 Elsevier B.V., All rights reserved.