Kocaeli University Research Information System
Not a member yet
80615 research outputs found
Sort by
Deep learning model for automated segmentation of sphenoid sinus and middle skull base structures in CBCT volumes using nnU-Net v2
No full textObjective: The purpose of this study is the development of a deep learning model based on nnU-Net v2 for the automated segmentation of sphenoid sinus and middle skull base anatomic structures in cone-beam computed tomography (CBCT) volumes, followed by an evaluation of the model’s performance. Material and methods: In this retrospective study, the sphenoid sinus and surrounding anatomical structures in 99 CBCT scans were annotated using web-based labeling software. Model training was conducted using the nnU-Net v2 deep learning model with a learning rate of 0.01 for 1000 epochs. The performance of the model in automatically segmenting these anatomical structures in CBCT scans was evaluated using a series of metrics, including accuracy, precision, recall, dice coefficient (DC), 95% Hausdorff distance (95% HD), intersection on union (IoU), and AUC. Results: The developed deep learning model demonstrated a high level of success in segmenting sphenoid sinus, foramen rotundum, and Vidian canal. Upon evaluation of the DC values, it was observed that the model demonstrated the highest degree of ability to segment the sphenoid sinus, with a DC value of 0.96. Conclusion: The nnU-Net v2-based deep learning model achieved high segmentation performance for the sphenoid sinus, foramen rotundum, and Vidian canal within the middle skull base, with the highest DC observed for the sphenoid sinus (DC: 0.96). However, the model demonstrated limited performance in segmenting other foramina of the middle skull base, indicating the need for further optimization for these structures
Thermophysical Properties of TiO2-Based Mono and Hybrid Nanofluids: Impact of CuO, ZnO, and Al2O3 Additives on Thermal Conductivity and Viscosity
No full textThis study investigates the thermophysical properties of mono and hybrid nanofluids based on TiO2 nanoparticles dispersed in bidistilled water (DDW), with the addition of CuO, ZnO, and Al2O3 at a total volume concentration of 4 %vol in a 1:1 ratio. Nanofluids were synthesized using a two-step method with ultrasonic dispersion and surfactant stabilization (SDBS, 1:0.1 ratio). Thermal conductivity was measured using the transient hot wire method, while kinematic viscosity was assessed across a temperature range of 293 K to 333 K. Results showed that all nanofluids exhibited improved thermal conductivity and higher viscosity compared to pure DDW. Among them, the TiO2–CuO/DDW hybrid demonstrated the best overall performance, with a thermal conductivity increase of up to 14 % and the lowest relative increase in viscosity. In contrast, TiO2–Al2O3/DDW showed the highest viscosity increase (up to 140 % at 293 K) and the lowest conductivity enhancement. Additionally, the experimental thermal conductivity data were compared with theoretical models, revealing that the Maxwell model consistently showed the closest agreement, with minimal deviations across all nanofluids (e.g., MAPE: 1.1 % for TiO2 and 1.2 % for TiO2–ZnO). In terms of viscosity modeling, the Maïga model provided the most accurate predictions in most cases, particularly for TiO2–CuO (MAPE: 4.3 %), while the Pak-Cho model significantly overestimated viscosity in hybrid nanofluids, with errors exceeding 100 %. These findings suggest that CuO and ZnO nanoparticles are more effective than Al2O3 in improving heat transfer while minimizing flow resistance, making it better suited for practical thermal applications
Sonlu Elemanlar Yöntemini Kullanarak HDPE Şişe Üretiminde Parizon Kontrolü Ve Ağırlık Optimizasyonu İle Enerji Ve Çevre Analizi
No full textEffectiveness of classification and scoring systems in solitary fibrous tumor prognosis prediction
No full textKRİTİK ACİL SERVİS HASTALARINDA SIVI TEDAVİSİNİN DEĞERLENDİRİLMESİNDE SOL VENTRİKÜL ÇIKIŞ YOLU HIZZAMAN İNTEGRALİ İLE KAROTİS ARTER HIZ-ZAMAN İNTEGRALİ VE DÜZELTİLMİŞ KAROTİS AKIM ZAMANI ARASINDAKİ KORELASYONUN ARAŞTIRILMASI
No full textVoxVeritasNet: A new feature engineering model leveraging iterative feature selection for detecting fake or real speech
No full textThis study introduces VoxVeritasNet, a high-precision and computationally efficient feature engineering framework for deepfake audio detection. The methodology leverages a nine-level Multi-Level Discrete Wavelet Transform (MDWT) to capture intricate time-frequency artifacts. A key innovation is the quantum-inspired dual-path mapping algorithm, which models parallel signal dependencies and embeds features into a high-dimensional Hilbert space for enhancing geometric separability. To optimize performance, an iterative ensemble selection strategy utilizing Neighborhood Component Analysis (NCA), Chi2, and ReliefF is employed alongside Support Vector Machines and k-Nearest Neighbors. The framework was evaluated across three public datasets with varying class distributions, achieving state-of-the-art peak accuracies of 99.96% with db4 and 99.71% with sym8 wavelets. Even using with the computationally efficient sym4 baseline, the model maintained exceptional detection rates above 98.99% and an equal error rate (EER) as low as 0.14%. VoxVeritasNet operates with a processing throughput of 6.45 segments per second on standard CPU hardware with a negligible storage footprint, offering a lightweight and explainable alternative to resource-intensive deep learning architectures
Comparison of procalcitonin, C-reactive protein and neutrophil/lymphocyte ratio in prediction of noninvasive mechanical ventilation failure in patients admitted to the emergency department with COPD exacerbation
No full textBackground: Non-invasive mechanical ventilation (NIMV) represents a cornerstone therapy for acute chronic obstructive pulmonary disease (COPD) exacerbation in emergency department (ED) settings. Objectives: Clinical predictors of NIMV response remain poorly characterized. This study sought to evaluate and compare the predictive capacity of different biomarkers for identifying patients at risk of NIMV failure during acute exacerbations. Methods: This prospective cohort study was conducted in the ED of a tertiary center from March 2023 to December 2024. Consecutive patients presenting with acute COPD exacerbations and meeting criteria for NIMV were enrolled. The primary outcome (NIMV failure) was evaluated during the initial 2-h monitoring period. The predictive performance of C-reactive protein (CRP) and procalcitonin levels, and neutrophil-to-lymphocyte ratio (NLR) were assessed using receiver operating characteristic (ROC) curve analysis. Areas under the curve (AUC) were compared using the de-Long method. Results: Among 151 enrolled patients, 73 (48.3 %) experienced NIMV failure, with an associated mortality rate of 30.1 %. For NIMV failure, ROC analysis demonstrated superior predictive performance for NLR (AUC = 0.804, 95 % confidence interval [CI]:0.734–0.875) compared to CRP (AUC = 0.680, 95 % CI:0.594–0.765) and procalcitonin (AUC = 0.682, 95 % CI:0.596–0.767). ROC analysis identified an optimal NLR cutoff of 5.8 for predicting NIMV failure, demonstrating 79.5 % sensitivity and 70.5 % specificity. When integrated with the HACOR score, this NLR threshold showed enhanced specificity with reduced sensitivity. Conclusion: The present study demonstrated that NLR was the strongest predictor of NIMV failure in the ED compared to CRP or procalcitonin. The combination of biomarkers with the HACOR score significantly enhanced prognostic accuracy