Atılım Academic Archive (Atılım University)
Not a member yet
9347 research outputs found
Sort by
Performance Assessment of Anion Exchange Electrolyzer With PBI-BASED Membrane Through 0-D Modeling
No full textDevrim, Yilser/0000-0001-8430-0702Anion exchange membrane (AEM) water electrolysis is emerging as a promising method for the sustainable production of hydrogen. A key advantage lies in the potential for cost-effective hydrogen production by substituting expensive noble metal electrocatalysts with affordable transition metals. This work presents a 0-D mathematical model for evaluating the performance of AEMWEs, with a particular focus on polybenzimidazole (PBI)-based membranes, which are renowned for their high thermal stability, chemical resistance and excellent conductivity in alkaline media. The objective of the model is to predict the behavior of membranes in AEMWE systems, and it has been employed to evaluate the performance of a range of PBI membranes. To ensure precision, the values were meticulously selected from the literature, in accordance with the experimental conditions. Furthermore, IR-corrected validation was incorporated to isolate the impact of membrane conductivity on performance, thereby facilitating a dependable assessment of PBI membranes under diverse conditions. The model considers the effects of electrolyte resistance and bubble formation on cell voltage behavior. The efficiency was evaluated on the basis of the higher heating value (HHV). The findings demonstrate that one membrane exhibits consistent efficiency across a broad temperature range (40-90 degrees C), whereas the other displays notable variability under diverse conditions. In particular, the efficiency of the electrolyzer is significantly enhanced by the use of thinner membranes and higher temperatures. The highest efficiencies obtained were 83.9% and 79.8% for 25 mu m and 50 mu m PBI/Polystyrene membrane under the operating conditions of 1 M KOH solution at 80 degrees C and current density of 2 A/cm2. This study aims to provide valuable information on the performance of PBI membranes through a zero-dimensional model validated by experimental data.Scientific and Technological Research Council of Turkiye (TUBITAK) [123M878]; TUBITAKThis study was supported by Scientific and Technological Research Council of Turkiye (TUBITAK) under Grant Number 123M878. The authors thank TUBITAK for their support.Science Citation Index Expande
Molecular/Antigenic Mimicry and Immunological Cross-Reactivity Explains Sars-Cov Autoimmunity
No full textCOVID-19 pandemic is over, but its effects on chronic illnesses remain a challenging issue. Understanding the influence of SARS-COV-2-mediated autoimmunity and overt autoimmune disease is of paramount importance, as it can provide a critical mass of information regarding both infection-mediated (and vaccination-induced) autoimmune phenomena in susceptible individuals during the disease course, and short or long-term post-disease sequelae. The high prevalence of organ and non-organ specific autoantibody positivity in patients with COVID-19 led to studies attempting to delineate the origin and the underlying mechanism responsible for their induction nature, identifying novel autoantigens, and the self-epitope sequences which could be the impetus for the initiating autoreactive responses. Herein, we provide a meticulous review of the studies reporting those mimicking sequences that have been experimentally validated, based on the assumption that molecular mimicry and immunological crossreactivity may account for autoantibody development. Most reports are based on bioinformatics approaches, and only a disproportionally small number of studies currently demonstrate immunological crossreactivity. We took the opportunity to further review and searched for the linear human epitope sequences of human, through the epitopes deposited at the Immune Epitope Database. This included an analysis of autoimmune disease as the disease data to comprehensively understand the subject matter. The critical overview of the findings underscore the urgent and immense need for further research to gain a comprehensive understanding of the mechanisms involved and the anticipated appraisal that molecular mimicry and immunological crossreactivity is indeed central to the loss of immunological tolerance during SARS-COV-2 infection.Science Citation Index Expande
The Role of a Smartphone Application in Monitoring the Risk of Hearing Loss Associated With Personal Listening Devices in Young Adults
No full textBackground Exposure to loud music has been reported to affect high-frequency hearing thresholds in adults. This study aimed to use a mobile application to assess the risk of hearing loss associated with personal listening devices (PLDs) in young adults. Methods A cross-sectional study was conducted on 99 healthy iPhone Operating System (iOS) smartphone users aged 17 to 31 years. Based on their weekly, monthly, 6-month, and annual listening data from the "Health" application, participants were classified into at-risk or non-risk groups. Six individuals in the at-risk group underwent audiological assessments. The Speech, Spatial, and Qualities of Hearing Scale (SSQ) was given to all participants. SSQ scores were compared between groups using the Mann-Whitney U test. The correlation between average PLD sound levels and SSQ scores was analyzed using the Spearman's test. Results The weekly, monthly, 6-month, and annual listening data showed that 16, 14, 12, and 12 participants were at risk, respectively. All six participants who underwent audiological assessment exhibited normal hearing thresholds. However, a pattern of decline at high frequencies was noted. A significant difference was found between the groups in terms of overall SSQ score (z = -2.74, P = 0.00), speech perception score (z = -3.25, P = 0.00), and hearing quality score (z = -2.01, P = 0.04) based on the 6-month and annual listening data. A weak positive correlation was found between the SSQ scores across all subscales and monthly sound-level data. Exposure duration was negatively correlated with speech perception (r = -0.32, P < 0.05), spatial perception (r = -0.26, P < 0.05), and overall score (r = -0.29, P < 0.05) in weekly data, as well as with hearing quality (r = -0.21; P < 0.05) and overall score (r = -0.21, P < 0.05) in annual data. Conclusions The immediate effects of loud music exposure were absent; however, prolonged exposure resulted in reduced speech perception and hearing-quality levels. Data from the iOS "Health" application indicated that some individuals were at risk of hearing loss, suggesting the need to modify their listening habits to prevent long-term decline in hearing function.TUBITAK (The Scientific and Technological Research Council of Turkiye) [2209-A]Supported by TUBITAK (The Scientific and Technological Research Council of Turkiye) 2209-A University Students Research Projects Support Program.Science Citation Index Expande
Mold Growth Affecting the Achievement of NZEB in the Long Term in Tropical Climates
No full textThe net-zero energy concept significantly impacts global goals regarding energy accessibility (SDG 7) and responsible consumption (SDG 12), particularly in the building sector, which accounts for substantial energy use and greenhouse gas emissions. While extensive research on Net Zero Energy Buildings (NZEB) has focused on the global north, tropical regions require further study, where high solar radiation, temperatures, and humidity challenge building performance throughout the year. Addressing problems like mold growth caused by these tropical climate aspects can undermine NZEB's performance. This study aims to evaluate the impact of mold growth on a representative building under the tropical climate of Panama City (high temperatures and humidity) and Boquete (low temperatures and high humidity). Long-term hygrothermal and energy performance analyses are conducted using simulation software to assess when and how mold growth affects building performance. Mold can harm the health of occupants and increase energy consumption, as additional humidity control devices may be required after the building's design phase. © 2025 Elsevier B.V., All rights reserved
Factors Affecting Dentists' Intention To Adopt Artificial Intelligence: An Extension of the Unified Theory of Acceptance and Use of Technology (UTAUT) Model
Get PDFPurposeAdvancements in science and technology have integrated artificial intelligence (AI) into dentistry, improving treatment processes, operational efficiency, and clinical outcomes. However, AI adoption among dentists remains underexplored, hindering progress in oral healthcare. This study aims to identify key barriers to AI adoption and examine factors influencing dentists' intention to use AI.Design/methodology/approachA quantitative cross-sectional approach was employed, utilizing self-administered questionnaires distributed online and across various dental clinics and hospitals in Ankara, Turkey. A total of 440 dentists participated in the study. Data analysis was conducted using SPSS and SmartPLS.FindingsThe study found that AI-anxiety negatively affects the intention to adopt AI in dentistry, showing a medium (almost large) effect that is stronger than other UTAUT factors such as performance expectancy, effort expectancy, and social influence, which demonstrated only small effects. Dentists with higher anxiety about learning and sociotechnical blindness are less likely to adopt AI, while concerns about job replacement and AI-configuration have less but still significant impact.Research limitations/implicationsThese results contribute to the growing body of knowledge on technology adoption in oral healthcare and provide practical implications for technology developers, policymakers, and other stakeholders seeking to facilitate AI integration in dentistry.Originality/valueThis study provides novel insights into AI adoption in dentistry, offering guidance for future development and integration, and addressing a critical research gap in a growing field-particularly in Turkey, where implementation is still in its early stages
A Stochastic Programming Framework for Pricing and Market Share Optimization in Retail Systems
Get PDFThis study examines a scenario where a manufacturer owns one retailer and collaborates with an independent retailer to sell a single product with Poisson demand over a multi-period selling horizon. The manufacturer protects the independent retailer's profitability through price protection and mid-life and end-of-life return opportunities. The retailers are allowed to place replenishment orders throughout the selling horizon. The manufacturer-controlled and independent retailers manage their stocks through order-up-to policies and hybrid policies comprising order-up-to and dispose-down-to policies, respectively. We employ stochastic programming techniques to construct models to determine the manufacturer's optimal pricing strategy. Retail Fixed Markdown (RFM) policy is assumed to determine the retail price at which the independent retailer sells the product. We also consider the impact of retail prices on the retailers’ market shares, which influence the mean demand observed by each retailer. We propose a modified version of the Stochastic Dual Dynamic Programming (SDDP) algorithm to determine the manufacturer's approximately optimal pricing strategy. Then, we examine how price protection contract parameters affect the manufacturer's approximately optimal pricing strategy and the retailers’ expected total profits. We also make comments on the selection of ideal values for the parameters. © 2025 The Author(s)Gulf University for Science and Technology, GUS
Pankreatikoduodenektomi Sonrası Gelişen Postoperatif Pankreatik Fistül İçin Erken Bir Belirteç Olarak C-reaktif Protein
No full textAmaç: Pankreatikoduodenektomi (PD) sonrası gelişen klinik olarak anlamlı postoperatif pankreatik fistül (CR-POPF), cerrahi sonrası morbiditenin en önemli nedenlerinden biri olup, sekonder komplikasyonlara da zemin hazırlamaktadır. CR-POPF’nin erken dönemde tanınması, zamanında müdahale ve uygun dren yönetimi açısından kritik öneme sahiptir. Bu çalışmanın amacı, postoperatif 3. gün (POD 3) serum C-reaktif protein (CRP) düzeylerinin CR-POPF gelişimini öngörmedeki değerini değerlendirmektir. Gereç ve Yöntem: Mart 2020 ile Şubat 2025 arasında merkezimizde PD uygulanan 112 hasta retrospektif olarak analiz edildi. Total veya distal pankreatektomi yapılanlar, neoadjuvan tedavi alanlar, kontrolsüz diyabeti olanlar, immünsüprese hastalar ve pankreas dışı anastomoz kaçağı gelişenler çalışma dışı bırakıldı. Postoperatif 3. gün serum CRP düzeyleri, dren amilaz konsantrasyonu ve drenaj hacmi kaydedildi. CR-POPF, Uluslararası Pankreas Cerrahisi Çalışma Grubu (ISGPS) kriterlerine göre tanımlandı. ROC eğrisi analizi ile CRP düzeylerinin tanısal performansı değerlendirildi. Ayrıca, çok değişkenli lojistik regresyon analizi ile CR-POPF için bağımsız risk faktörleri belirlendi. Bulgular: CR-POPF, 17 hastada (%15,2) gelişti. Bu hastalarda pankreas duktus çapı anlamlı şekilde daha dar ve doku yapısı daha yumuşaktı (p 161 mg/L olan hastalarda CR-POPF gelişimini öngörmede AUC 0,77, sensitivite %82,4 ve spesifisite %66,3 olarak bulundu. Lojistik regresyon analizinde, dar pankreatik duktus çapı ve yumuşak pankreas dokusu CR-POPF için bağımsız risk faktörleri olarak saptandı. Sonuç: Pankreatikoduodenektomi sonrası 3. gün serum CRP düzeylerinin yüksek olması, CR-POPF gelişimini öngörmede anlamlı bir belirteçtir. CRP düzeylerinin rutin postoperatif değerlendirmeye dahil edilmesi, erken risk sınıflamasını kolaylaştırarak hasta yönetimi ve dren çekilme zamanlamasında klinik karar sürecine katkı sağlayabilir
Assessment of Anticancer Effects of Aloe Vera on 3D Liver Tumor Spheroids in a Microfluidic Platform
No full textTevlek, Atakan/0000-0003-0601-8642The search for effective anticancer therapies has increasingly focused on natural compounds like Aloe vera, renowned for its therapeutic properties. This study investigates the anticancer properties of Aloe vera on 3D liver tumor spheroids via a PDMS-based microfluidic device, providing a more physiologically realistic model compared to traditional 2D cultures. HepG2 cells were cultivated to generate 3D spheroids on-chip, thereafter subjected to different concentrations of Aloe vera and the chemotherapeutic drug Doxorubicin to evaluate cytotoxic effects. The microfluidic system, validated by COMSOL simulations, facilitated continuous perfusion and real-time assessment of cell viability over a duration of 10 days. The results indicated that Aloe vera markedly diminished cell viability by triggering apoptosis at concentrations over 12.5 mg/mL. IC50 values were determined at 72 h: 25 +/- 0.10 mg/mL for Aloe vera and 5.47 +/- 0.03 mu g/mL for Doxorubicin in 2D cultures, but in 3D cultures, the IC50 values were 31.25 +/- 0.14 mg/mL for Aloe vera and 8.33 +/- 0.05 mu g/mL for Doxorubicin. This study underscores the promise of Aloe vera as a natural anticancer agent and illustrates the efficacy of microfluidic platforms for enhanced drug screening and customized medicine applications.Turkiye Bilimsel ve Teknolojik Arastirma KurumuThe study was supported by the Turkiye Bilimsel ve Teknolojik Arastirma Kurumu.Science Citation Index Expande
A Modular Multi-Stage Method for Vehicle Detection and Classification in Low Resolution Images
No full textAkıllı ulaşım sistemlerinde (ITS) gerçek zamanlı araç tespitinin önemi, şehir trafiğindeki araç sayısındaki sonsuz ve sürekli artışla vurgulanmaktadır. Bununla birlikte, çok çeşitli kamera kaliteleri ve çözünürlükleri, farklı görüş açıları ve zayıf aydınlatma ve olumsuz hava koşulları gibi harici ve kontrol edilemeyen değişkenlerin etkisi, doğru araç tespiti ve sınıflandırmasında birçok zorluk yaratmaktadır. Derin öğrenme tabanlı nesne algılama algoritmalarının çoğu, daha önce bahsedilen bu koşullar düşük görünürlük ve/veya düşük çözünürlüklü görüntülere neden olduğu için bu tür durumlarda zorlanmaktadır. Bu kısıtlamaların üstesinden gelmek için bu çalışma, loş ışık, kötü hava koşulları ve düşük çözünürlük gibi zorlu görüntüleme durumlarına uyarlanmış gerçek zamanlı araç tespiti ve sınıflandırması için yeni, modüler, etkili ve güvenilir bir yaklaşım önermektedir. Önerilen yaklaşım iki özel veri kümesinin oluşturulmasını içermektedir. İlk veri kümesi PASCAL VOC formatında 4.500 düşük çözünürlüklü trafik manzarası görüntüsünden oluşmakta ve transfer öğrenme yoluyla bir nesne tespit modelini eğitmek için kullanılmaktadır. İkinci veri kümesi, iki farklı sınıflandırma modelini eğitmeyi amaçlayan, her biri 100 × 100 piksel boyutlarında ve 96 dpi ve altında çözünürlüğe sahip beş araç türünün 10.000 düşük çözünürlüklü görüntüsünü içerir. Önerilen yaklaşım, son teknoloji ürünü tek aşamalı bir dedektör (SSD) olan EFFICIENTDET1'i hafif bir özel evrişimli sinir ağı (CNN) sınıflandırıcısı ve bir XGBoost sınıflandırıcısı ile entegre etmektedir. Bu kombinasyon, hem makine hem de derin öğrenme algoritmalarının güçlü yönlerinden faydalanarak tespit performansını ve sınıflandırma doğruluğunu artırır. Önerilen yaklaşımın etkinliği deneysel değerlendirme ile gösterilmiştir. Önerilen yaklaşım, 0,9323 ortalama ortalama hassasiyet (mAP) ile aynı veri kümesi üzerinde karşılaştırılabilir koşullarda geleneksel ve son teknoloji nesne algılama modellerinden belirgin şekilde daha iyi performans göstermektedir. Ayrıca, çoklu işlemin uygulandığı önerilen yaklaşım, kare başına 26 milisaniyelik bir çıkarım hızına ulaşmaktadır. Bu, son teknoloji ürünü nesne yöntemlerine kıyasla hem doğruluk hem de çıkarım hızında önemli bir gelişmeye işaret etmektedir. Önerilen yaklaşımın modüler, uyarlanabilir ve ölçeklenebilir yapısı, onu ITS'deki uygulamalar için ideal kılmaktadır. Önerilen yaklaşımın yüksek doğruluğunun yanı sıra çıkarım hızı, düşük görüntü kalitesi veya olumsuz çevresel faktörler gibi koşullar altında gerçek zamanlı uygulamalar için etkili ve operasyonel bir seçenek haline getirmektedir. Sonuç olarak, önerilen yaklaşım, zorlu durumlarda daha güvenli ve daha etkili ulaşım yönetimi sağlayabileceğinden, derin öğrenme tabanlı araç algılama alanında büyük bir potansiyele sahiptir. Bu bulgular, verimli bir nesne algılama modelinin çok işlemli bir mimaride özel sınıflandırıcılarla birleştirilmesinin, gerçek zamanlı araç algılamada gelecekteki araştırmalar için umut verici bir yönü temsil ettiğini göstermektedir.The relevance of real-time vehicle detection in intelligent transportation systems (ITS) is highlighted by the endless and continuous rise in the number of vehicles in the urban traffic. However, the wide range of camera qualities and resolutions, different viewing angles, and the impact of external and uncontrollable variables like poor lighting and adverse weather create many difficulties in accurate vehicle detection and classification. The majority of deep learning-based object detection algorithms struggles in such circumstances as these before-mentioned conditions cause low-visibility and/or low-resolution images. In order to overcome these constraints, this study suggests a novel, modular, effective and reliable approach for real-time vehicle detection and classification adapted to difficult imaging situations, such as dim lighting, bad weather, and low-resolution situations. The proposed approach involves the creation of two custom datasets. The first dataset comprises 4,500 low-resolution traffic scenery images formatted in PASCAL VOC and is used to train an object detection model through transfer learning. The second dataset includes 10,000 low-resolution images of five types of vehicles, each with dimensions of 100 × 100 pixels and a resolution of 96 dpi (dot per inch) and below, aimed at training two different classification models. The proposed approach integrates EFFICIENTDET1, a state-of-the-art single-stage detector (SSD), with a lightweight custom convolutional neural network (CNN) classifier and an XGBoost classifier. This combination improves detection performance and classification accuracy by taking advantages of both machine and deep learning algorithms strengths. The effectiveness of the proposed approach is demonstrated by the experimental evaluation. With a mean average precision (mAP) of 0.9323, the proposed approach performs noticeably better than traditional and state-of-the-art object detection models in comparable circumstances on the same dataset. Additionally, the proposed approach, in which multiprocessing has been implemented, achieves an inference speed of 26 milliseconds per frame. This marks a substantial improvement in both accuracy and inference speed compared to state-of-the-art object methods. The modular, adaptable, and scalable nature of the proposed approach makes it ideal for applications in ITS. The inference speed along with the high accuracy of the proposed approach make it effective and an operational option for real-time applications under conditions such as low image quality or adverse environmental factors. Consequently, the proposed approach has plenty of potential in deep learning based vehicle detection area as it can enable safer and more effective transportation management in challenging situations. These findings suggest that combining an efficient object detection model with custom classifiers in a multiprocessing architecture, represents a promising direction for future research in real-time vehicle detection
Minimization of Greenhouse Effects by Optimal Plankton Feeding: A Simulation-Based Study
No full textGlobal warming and related greenhouse effects possess significant threats to environmental sustainability. This research investigates the possibility of reducing the greenhouse gas levels and associated ambient temperature by manipulating the plankton population in a given forecasting period. To achieve this goal, an optimal control strategy is developed by Pontryagin’s minimum principle, and it is applied to a recently derived nonlinear marine ecosystem model describing the variation of greenhouse gas levels, ambient temperature, and fish interactions. The main goal is to determine an external plankton generation profile that is expected to reduce the greenhouse gas levels and associated ambient temperature to the highest possible extent. The simulation results reveal that the optimal feeding strategy enables one to achieve a reduction of 54% in greenhouse gas levels and 95% in the associated ambient temperature. This research proposes a biological-based novel control approach that can serve as an alternative solution to environmental degradation. © 2025 Elsevier B.V., All rights reserved