Altınbaş University Institutional Repository
Not a member yet
5805 research outputs found
Sort by
Optimal feature tuning model by variants of convolutional neural network with LSTM for driver distract detection in IoT platform
No full textNowadays, traffic accidents are caused due to the distracted behaviors of drivers that have been noticed with the emergence of smartphones. Due to distracted drivers, more accidents have been reported in recent years. Therefore, there is a need to recognize whether the driver is in a distracted driving state, so essential alerts can be given to the driver to avoid possible safety risks. For supporting safe driving, several approaches for identifying distraction have been suggested based on specific gaze behavior and driving contexts. Thus, in this paper, a new Internet of Things (IoT)-assisted driver distraction detection model is suggested. Initially, the images from IoT devices are gathered for feature tuning. The set of convolutional neural network (CNN) methods like ResNet, LeNet, VGG 16, AlexNet GoogleNet, Inception-ResNet, DenseNet, Xception, and mobilenet are used, in which the best model is selected using Self Adaptive Grass Fibrous Root Optimization (SA-GFRO) algorithm. The optimal feature tuning CNN model processes the input images for obtaining the optimal features. These optimal features are fed into the long short-term memory (LSTM) for getting the classified distraction behaviors of the drivers. From the validation of the outcomes, the accuracy of the proposed technique is 95.89%. Accordingly, the accuracy of the existing techniques like SMO-LSTM, PSO-LSTM, JA-LSTM, and GFRO-LSTM is attained as 92.62%, 91.08%, 90.99%, and 89.87%, respectively, for dataset 1. Thus, the suggested model achieves better classification accuracy while detecting distracted behaviors of drivers and this model can support the drivers to continue with safe driving habits
De-escalation of nodal surgery in clinically node-positive breast cancer
No full textImportance: Increasing evidence supports the oncologic safety of de-escalating axillary surgery for patients with breast cancer after neoadjuvant chemotherapy (NAC).
Objective: To evaluate the oncologic outcomes of de-escalating axillary surgery among patients with clinically node (cN)-positive breast cancer and patients whose disease became cN negative after NAC (ycN negative).
Design, setting, and participants: In the NEOSENTITURK MF-1803 prospective cohort registry trial, patients from 37 centers with cT1-4N1-3M0 disease treated with sentinel lymph node biopsy (SLNB) or targeted axillary dissection (TAD) alone or with ypN-negative or ypN-positive disease after NAC were recruited between February 15, 2019, and January 1, 2023, and evaluated.
Exposure: Treatment with SLNB or TAD after NAC.
Main outcomes and measures: The primary aim of the study was axillary, locoregional, or distant recurrence rates; disease-free survival; and disease-specific survival. Number of axillary lymph nodes removed was also evaluated.
Results: A total of 976 patients (median age, 46 years [range, 21-80 years]) with cT1-4N1-3M0 disease underwent SLNB (n = 620) or TAD alone (n = 356). Most of the cohort had a mapping procedure with blue dye alone (645 [66.1%]) with (n = 177) or without (n = 468) TAD. Overall, no difference was found between patients treated with TAD and patients treated with SLNB in the median number of total lymph nodes removed (TAD, 4 [3-6] vs SLNB, 4 [3-6]; P = .09). Among patients with ypN-positive disease, those who underwent TAD were more likely to have a lower median lymph node ratio (TAD, 0.28 [IQR, 0.20-0.40] vs SLNB, 0.33 [IQR, 0.20-0.50]; P = .03). At a median follow-up of 39 months (IQR, 29-48 months), no significant difference was found in the rates of ipsilateral axillary recurrence (0.3% [1 of 356] vs 0.3% [2 of 620]; P ≥ .99) or locoregional recurrence (0.6% [2 of 356] vs 1.1% [7 of 620]; P = .50) between the TAD and SLNB groups, with an overall locoregional recurrence rate of 0.9% (9 of 976). The initial clinical tumor stage, pathologic complete response, and use of blue dye alone as a mapping procedure were not associated with the outcome. Even though patients with TAD demonstrated an increased disease-free survival rate compared with the SLNB group, this difference did not reach statistical significance (94.9% vs 92.6%; P = .07). Factors associated with decreased 5-year disease-specific survival were cN2-3 axillary stage (cN1, 98.7% vs cN2-3, 96.8%; P = .03) and nonluminal type tumor pathologic characteristics (luminal, 98.9% vs nonluminal, 96.9%; P = .007).
Conclusions and relevance: The short-term results suggest very low rates of axillary and locoregional recurrence in a select group of patients with cN-negative disease after NAC treated with TAD alone or SLNB alone followed by regional nodal irradiation regardless of the SLNB technique or nodal pathology. Whether TAD might provide a clear survival advantage compared with SLNB remains to be proven in studies with longer follow-up
Investigation of the tissue equivalence of typical 3D-printing materials for application in internal dosimetry using monte carlo simulations
Full text linkThis study evaluates the dosimetric accuracy of PLA and ABS 3D-printed phantoms compared to real tissues using Monte Carlo simulations in radionuclide therapy.
Materials and methods: A phantom representing average liver and lung volumes, with a 10 mm tumor mimic in the liver, was simulated for radioembolization using 1 mCi Tc-99 m and 1 mCi Y-90. The dose distribution (DD) was compared across PLA, ABS, and real organ densities.
Results: For Tc-99 m, PLA showed a + 5.6% DD difference in the liver, and ABS showed - 35.3% and - 40.9% differences in the lungs. For Y-90, PLA had a + 1.7% DD difference in the liver, while ABS showed - 34.2% and - 34.9% differences in the lungs.
Conclusion: In MC simulation, PLA is suitable for representing high-density tissues, while ABS is appropriate for simulating moderately low-density tissues
Assessment of historic civil buildings in Istanbul for their sustainable conservation
No full textPurposeSimilar to monumental historic structures, civil historic buildings should also be paid great attention since they form the historic environment and reflect the historically dynamic nature of a city. However, these buildings are gradually disappearing in Istanbul. This study aims to examine the current situation of civil historic buildings in Istanbul and determine the reasons for their gradual disappearance.Design/methodology/approachTwo extensive site surveys have been performed in the & Uuml;sk & uuml;dar district of Istanbul. The first survey was based on the visual inspection of the buildings, while the second one contained detailed questionnaires performed with the owners of the historic buildings.FindingsThe performed surveys revealed the conservation status of the examined historical buildings. The details on their structural systems, construction materials, initial and current functions were quantified. The contentment of the owners and financial support given by the governmental agencies were revealed. The effective sources of the damage seen in the studied historic buildings were also evaluated. Finally, the evaluation of the principles of sustainability showed that the adaptive reuse strategy has not been successfully implemented in Istanbul, and it can promote the conservation of historic buildings in & Uuml;sk & uuml;dar.Research limitations/implicationsThe first survey covered all historic civil buildings in & Uuml;sk & uuml;dar, while in total 206 of them out of 1,034 have been studied extensively in the second survey.Practical implicationsThe obtained results give an insight and/or understanding of historic civil buildings in Istanbul.Social implicationsThe results of the study are important for the conservation of historic buildings.Originality/valueThe performed study not only reveals the current situation of the historic civil buildings in & Uuml;sk & uuml;dar, Istanbul, but also analyses the attitudes of their owners. The effectiveness of the given governmental supports to the owners was also evaluated
Assessment of UAV Usage for Flexible Pavement Inspection Using GCPs: Case Study on Palestinian Urban Road
Full text linkArticle number : 8129Rehabilitation plans are based on pavement condition assessments, which are crucial to modern pavement management systems. However, some of the disadvantages of conventional approaches for road maintenance and repair include the time consumption, high costs, visual errors, seasonal limitations, and low accuracy. Continuous and efficient pavement monitoring is essential, necessitating reliable equipment that can function in a variety of weather and traffic conditions. UAVs offer a practical and eco-friendly alternative for tasks including road inspections, dam monitoring, and the production of 3D ground models and orthophotos. They are more affordable, accessible, and safe than traditional field surveys, and they reduce the environmental effects of pavement management by using less fuel and producing less greenhouse gas emissions. This study uses UAV technology in conjunction with ground control points (GCPs) to assess the kind and amount of damage in flexible pavements. Vertical photogrammetric mapping was utilized to produce 3D road models, which were then processed and analyzed using Agisoft Photoscan (Metashape Professional (64 bit)) software. The sorts of fractures, patch areas, and rut depths on pavement surfaces may be accurately identified and measured thanks to this technique. When compared to field exams, the findings demonstrated an outstanding accuracy with errors of around 3.54 mm in the rut depth, 4.44 cm2 for patch and pothole areas, and a 96% accuracy rate in identifying cracked locations and crack varieties. This study demonstrates how adding GCPs may enhance the UAV image accuracy, particularly in challenging weather and traffic conditions, and promote sustainable pavement management strategies by lowering carbon emissions and resource consumption
Is there a lack of equity, diversity and inclusion in orthopaedic education? – A systematic review
No full textBackground: Trauma & Orthopaedics has failed to progress in promoting equity, diversity and inclusion. Only 7 % of orthopaedic consultants are female, and less than 1 % identify as black. Improved equity, diversity and inclusion within healthcare has shown to improve patient outcomes, innovation and reduce unconscious bias within the specialty. Patients within the NHS face significant disparities in the care they receive, and promoting EDI within orthopaedic education will help address this. Methods: A systematic literature search between 2020 and 2025 was performed to identify any literature. The electronic databases of Medline, EBSCO, Web of Science, PubMed and Scopus were utilized. Studies were screened against strict inclusion and exclusion criteria to ensure the literature was relevant and specific to this review. Two independent assessors read the abstracts and full-texts of articles to determine whether or not they were included within this systematic review. This systematic review was registered with PROSPERO (registration number: CRD420251022275) Results: The studies within this systematic review highlighted the absence of female and minority ethnic representation within the orthopaedic specialty. Additionally, several barriers to the implementation of EDI have been addressed. These include: a lack of female role models, a lack of commitment from staff, difficulties changing previous perceptions of the specialty and difficulty changing a lack of cultural sensitivity. Conclusion: Women and minority ethnic groups remain markedly underrepresented in orthopaedics, facing structural and cultural challenges that hinder recruitment and retention. Orthopaedic education plays a key role in preventing these issues, as medical students are often discouraged from pursuing this specialty from the outset. Misconceptions and the absence of diverse role models are key factors in deterring students from under-represented backgrounds from entering the specialty. Promoting a culture of inclusion will help change these perceptions and misconceptions. This will aid in enhanced patient-centred care and better healthcare outcomes for all. Despite growing recognition and awareness of the limited diversity in orthopaedics, there remains a shortage of literature that collates data together. This review does this by critically analysing existing research and offers a comprehensive review of this topic
A hybrid firefly-Bayesian Neural Network approach for survival and reliability estimation in non-homogeneous Poisson processes
Full text linkWe develop a new parametric inference scheme, which will combine Bayesian Neural Networks (BNNs) and Firefly Algorithm (FFA) in order to estimate the parameters of the Gull-Alpha Power Gompertz-Makeham (GAPL-GM) distribution fitted in a Non-Homogeneous Poisson Process (NHPP) under progressive Type-I censoring. This formulation of the problem of optimizing weight-prior parameters as a population-based metaheuristic issue works well to overcome non-convexity and multimodality of likelihood surfaces by the use of hybrid BNN-FFA. We consider weights of networks as random variables and so uncertainty can be quantified through the variational inference. The FFA is an optimization approach to the variational objective, the minimization of the negative log-likelihood with regularization, which finds optimal swarm of candidate weight vectors, where positive values of brightness reflect reduced values of loss. This mixed method provides strong parameter estimates of GAPL-GM model in situations of highly complicated censoring and small sample sizes. Simulated dataset and real-life failure time data are measurably evaluated by using the proposed method with fantastically better estimation accuracy based on the empirical evaluation. AIC, BIC, goodness-of-fit metrics (e.g. Anderson-Darling, Cramer-von Mises), and others show better fits of models compared with traditional maximum likelihood estimation (MLE) and ordinary optimization routines. Like others, the BNN-FFA method also offers credible provisions of the parameter estimators and the reliability functions that were expected as it was based on probabilistic rigor. To conclude, the study adds a highly competitive and uncertainty conscious estimation process on NHPP models in the situation of complicated lifetime regimes under censorship. The given BNN-FFA model has the flexible, interpretable, experimentally confirmed alternative to the classical estimating methods of analyzing statistical reliability and lifetime data
Enhanced forecasting of stock prices using long short-term memory networks: A comparative study with traditional models
No full textStock market prediction is an important research domain for formulating strategies that could maximize returns on investment. Traditional models, like ARIMA and ETS, often fail to take care of the linearity assumptions and miss capturing the complex, nonlinear trends of the stock prices. This paper proposes a model using Long Short-Term Memory (LSTM) networks for improving the quality of forecasts in terms of closing prices of the NIFTY50 index and benchmarking accuracy with models ARIMA and ETS. Using a dataset extracted from Yahoo Finance from January 1, 2008, to December 30, 2022, the results showed that an LSTM model was representative of an accuracy that excelled primarily in the longest forecasting horizons. RMSE and MAPE were among the observatory metrics used, all of which underscored the strength of LSTM in handling temporal dependencies and nonlinearities. This paper highlights the potential for state-of-the-art deep learning approaches to play important roles in financial forecasting and their implications in emerging markets for various investment strategies
Comparative analysis of autophagy and apoptosis pathways in viral and alcoholic cirrhosis: An immunohistochemical study
Full text linkObjectives: To study the different cellular death mechanisms between viral cirrhosis and alcoholic cirrhosis. The research investigated autophagy and apoptosis mechanisms in hepatitis B virus (HBV), hepatitis C virus (HCV) and alcohol-induced cirrhosis.
Methods: The research team analyzed biopsy samples from the liver which were obtained at Florence Nightingale Hospitals, Istanbul, Turkey. The experimental protocols were performed between February 2021 and February 2023. The study included 19 HBV, 15 HCV, 13 alcohol-related cirrhosis patients and 6 normal liver tissues. Beclin-1, Caspase-3, Bcl-2 and LC3 expressions were evaluated by immunohistochemistry.
Results: Staining intensity as well as extent underwent evaluation through H-score methodology. The expression levels of Beclin-1 (control: 0.7±0.3, HBV: 6.0±1.4, HCV: 5.1±1.3, alcohol: 4.8±1.2), caspase-3 (control: 0.4±0.2, HBV: 6.0±1.4, HCV: 5.1±1.2, alcohol: 5.5±1.3) and Bcl-2 (control: 0.3±0.2, HBV: 5.5±1.2, HCV: 4.8±1.1, alcohol: 4.6±1.1) were significantly higher than those of the control group (p<0.001). LC3 expression revealed no between-group differences. A positive association for Beclin-1 with Caspase-3 (r=0.582) alongside negative association for Bcl-2 with Caspase-3 (r=-0.608) was documented.
Conclusion: Our findings suggest that both autophagic and apoptotic pathways are active in the pathogenesis of cirrhosis. Similar cell death mechanisms were found to be involved in viral and alcoholic cirrhosis, but these pathways were more prominently activated in viral cirrhosis
Optimizing influenza prevention: a systematic review of the cost-effectiveness of pediatric vaccination programs and vaccine types
Full text linkIntroduction: Seasonal influenza is a major cause of illness and death worldwide. Vaccination remains the cornerstone of prevention, with options including trivalent inactivated (TIV), quadrivalent inactivated (QIV), and live-attenuated vaccines. This study aimed to provide a systematic overview of the cost-effectiveness of pediatric influenza vaccination programs, with a particular focus on comparing different vaccine types.
Methods: A comprehensive literature search was conducted in PubMed, Web of Science, Scopus, and Cochrane databases for records published between 2013 and 2024. The target population included individuals younger than 18 years. The primary research question was: Which influenza vaccines, trivalent, quadrivalent, or live-attenuated, are more cost-effective, and how does introducing seasonal vaccination for children under 18 influence healthcare costs and health outcomes? Data extraction was performed using a structured Excel spreadsheet.
Results: This review included 33 studies that met the inclusion and exclusion criteria. Most studies support the conclusion that vaccinating children is an effective and cost-effective strategy for reducing influenza transmission. Cost-effectiveness varied depending on epidemiological and demographic factors, the type of vaccine used, and age group differences, which were influenced by the analytical perspective and local health and economic conditions.
Conclusion: This review confirms that pediatric influenza vaccination is a cost-effective intervention, particularly with quadrivalent vaccines. The optimal choice of vaccine and strategy should be tailored to local population needs and economic conditions to maximize public health benefits