Aynur Yonar

Veterinary institutions need practical methods to translate the Sustainable Development Goals (SDGs) into profession-specific priorities. This study developed a two-phase priority-setting framework linking systematic SDG screening with multi-criteria decision analysis. First, all 169 SDG targets were screened using four veterinary-relevant dimensions. The 48 retained targets were synthesized into 27 preliminary concepts, refined through two rounds of structured internal expert evaluation, and harmonized by the research team into 27 final decision alternatives. Second, Budgetary Burden, Accessibility and Scope, Implementation Timeline, and Sustainability Impact were weighted using the Analytic Hierarchy Process (AHP), and the 27 alternatives were ranked using TOPSIS. Robustness was assessed through Monte Carlo criterion-weight perturbation, score aggregation, and leave-one-expert-out analyses. Budgetary Burden (0.4025) and Sustainability Impact (0.3526) received the largest weights. Prevention of Zoonotic Diseases ranked first (C_i = 0.8272), followed by Awareness-Raising Campaigns, Animal Welfare Inspections on Commercial Animal Farms, and Audits of Food Safety and Production Process Compliance Standards. Criterion-weight perturbation and score aggregation largely preserved the highest-priority set, whereas leave-one-expert-out analysis showed greater sensitivity to panel composition. The framework should therefore be interpreted as a transparent, context-dependent, priority-setting tool rather than as an externally validated or universally applicable hierarchy.

The Multivariate Gamma distribution is widely utilized in stochastic modeling, hydrology, reliability analysis, and actuarial sciences. Despite its broad applicability, efficient parameter estimation remains a challenge, as the maximum likelihood estimators lack closed-form solutions. Metaheuristic optimization techniques, particularly Particle Swarm Optimization (PSO), have been frequently employed to address this issue. However, the performance of PSO is highly sensitive to parameter settings, and in the absence of a structured approach, parameter tuning often relies on time-consuming trial-and-error methods. This study employs the Taguchi experimental design to systematically determine the optimal parameter configuration for PSO in estimating the parameters of the Multivariate Gamma distribution. By providing a structured framework for parameter tuning, the proposed approach reduces dependence on heuristic or arbitrary selection strategies. The performance of the Taguchi-optimized PSO is evaluated based on solution quality, measured using the deficiency (DEF) criterion, and computation time. The results indicate that the Taguchi method offers an efficient and replicable solution for PSO parameter tuning, eliminating the need for conventional trial-and-error methods. By structuring the tuning process, this approach enhances both the reliability and reproducibility of PSO-based estimation, making it adaptable to a range of similar optimization tasks. Future studies may explore comparative analyses with other parameter tuning strategies, extend the application to alternative metaheuristic algorithms, or adapt the methodology to optimization challenges in machine learning and data science contexts.

Diagnosing acute cholecystitis (AC) in patients with end-stage renal disease (ESRD) is challenging because chronic systemic inflammation and immune dysregulation may obscure acute inflammatory responses. Conventional inflammatory markers may therefore have limited diagnostic specificity in this population. This study aimed to evaluate the diagnostic utility of systemic inflammatory indices in ESRD patients undergoing cholecystectomy. This retrospective single-center study included 48 ESRD patients who underwent cholecystectomy either for clinically suspected AC (AC group, n = 31) or prophylactically prior to kidney transplantation (RT group, n = 17). Demographic characteristics and laboratory parameters were obtained from preoperative blood tests. Systemic inflammatory indices, including the neutrophil-to-lymphocyte ratio, systemic immune-inflammation index, systemic inflammation response index (SIRI), and pan-immune-inflammation value (PIV), were calculated. Histopathological findings were used as the reference standard for confirming acute inflammation. Exploratory receiver operating characteristic curve analysis was performed to evaluate diagnostic performance and determine optimal cutoff values. Histopathological examination demonstrated neutrophilic infiltration consistent with acute inflammation in 30% of patients with clinically suspected AC. Compared with the RT group, patients with histopathologically confirmed acute inflammation had significantly higher neutrophil counts and PIV levels and lower serum albumin levels. In receiver operating characteristic analysis comparing histopathologically confirmed acute cases with RT controls, PIV showed the highest diagnostic accuracy (area under the curve [AUC] = 0.771), followed by C-reactive protein (AUC = 0.733) and SIRI (AUC = 0.732). Within the AC cohort, SIRI demonstrated the best discriminatory ability for differentiating acute from chronic inflammatory changes (AUC = 0.732). Optimal cutoff values differed substantially between analyses, with lower thresholds observed when acute cases were compared with RT controls and higher thresholds required within the AC cohort. Composite inflammatory indices, particularly SIRI and PIV, may provide additional information in identifying histopathologically confirmed acute inflammation in ESRD patients. However, given the limited sample size, these findings should be interpreted as exploratory and require validation in larger prospective studies.

Reducing noise in medical imaging is essential to improve quality and diagnostic accuracy. Gaussian noise, common in MRI and X-ray images, reduces clarity and obscures pathological structures. Traditional denoising methods often fail to preserve edge details, leading to the loss of important anatomical information. To address this, an Adaptive Edge-Preserving Type-2 Fuzzy Filter is developed, aiming to reduce Gaussian noise while maintaining edge integrity. The proposed method integrates an adaptive mechanism into classical Type-2 Fuzzy Filtering, adjusting filtering parameters dynamically based on edge strength. This enables effective noise removal in both homogeneous and edge-rich regions. The performance of the methods was evaluated on real medical image datasets under varying noise levels and compared with Mean, Median, Fuzzy, and classical Type-2 Fuzzy Filtering methods. Results demonstrate superior performance in terms of Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM), providing a robust solution for improving medical image quality.

Artificial neural network (ANN) is one of the machine learning algorithms widely used in prediction studies recently. The key to obtaining effective prediction results with ANN depends on its training and the design of its tunable parameters. The Backpropagation and Levenberg-Marquardt (BP-LM) learning algorithms are the most utilized algorithms for training ANN. However, these algorithms have some disadvantages such as local minima, computational complexity, sensitivity to initialization, overfitting, and limited parallelism. In this study, we proposed a Particle Swarm Optimization (PSO)-trained ANN model to deal with these problems in ANN learning. PSO is one of the well-utilized artificial intelligence algorithms and it can be successful in the learning process thanks to its features of finding global optima, having a few parameters to be tuned, being easily parallelized, robustness and convergence speed. The proposed model is tested with different ANN structures and parameter values for tourism revenue prediction. As a result, it was observed that the proposed PSO-trained ANN model generally gave better prediction results than BP-LM trained ANN and an optimal ANN structure was obtained for the prediction of tourism revenues. In addition, forecasting of tourism revenues for the next 12 months was obtained with a designed optimal ANN structure.

To investigate novel inflammatory indexes including neutrophil‐to‐lymphocyte ratio (NLR), systemic immune inflammation index (SII), systemic immune response index (SIRI), and pan‐immune‐inflammation value (PIV) in the diagnosis of organizing pneumonia (OP) and distinguishing OP from lung cancer (LC). The study included 125 patients with LC and 21 patients with OP. Retrospectively, we compared the values of SII, SIRI, PIV, NLR, and C‐reactive protein (CRP) and evaluated correlations between the patient groups. The CRP levels in patients with OP (median: 45.2 mg/dL) were lower than those in patients with LC (median: 64 mg/dL). However, no differences were observed in systemic inflammatory indexes between the groups. Statistically significant positive correlations were found between neutrophils and NLR ( r = 0.63), SIRI ( r = 0.52), and SII ( r = 0.72). In addition, significant negative correlations were observed between lymphocytes and NLR ( r = −0.67), SIRI ( r = −0.53), SII ( r = −0.27), and PIV ( r = −0.37). Increased CRP levels may indicate lung carcinoma risk and CRP measurements may be predictive in distinguishing between OP and LC. Additionally, high neutrophil counts, thrombocyte counts and/or lymphocytopenia might lead to a poor immune response and are linked to higher systemic inflammatory indices.

Accurate automated classification of brain tumors from magnetic resonance imaging (MRI) is essential for early diagnosis and treatment. This study presents a hybrid framework combining Convolutional Neural Network (CNN) deep features, Large Margin Nearest Neighbor (LMNN) metric learning, and swarm-intelligence optimization for robust four-class classification. Five pretrained CNNs-DenseNet201, MobileNetV2, ResNet50, ResNet101, and InceptionV3-were evaluated on a dataset of 7,023 images categorized as glioma, meningioma, pituitary, healthy. Among these, DenseNet201 provided the highest baseline performance with 92.66% accuracy. LMNN improved feature separability, while Particle Swarm Optimization (PSO) and Grey Wolf Optimizer (GWO) selected compact subsets. The selected features were classified using k-Nearest Neighbor (KNN), Support Vector Machine (SVM), Artificial Neural Network (ANN), and Random Forest (RF). The DenseNet201-LMNN-GWO-KNN configuration, termed DenseWolf-K, achieved the best performance with 99.64% accuracy, establishing it as the optimal implementation of the framework. Robustness and generalizability were further confirmed using an independent external dataset. Model explainability was ensured through feature-level ranking of GWO-selected features and occlusion sensitivity maps, an Explainable Artifical Intelligence (XAI) method. Overall, the proposed DenseWolf-K framework delivers high accuracy, low false-negative rates, compact representation, and enhanced interpretability, representing a reliable and efficient solution for MRI-based brain tumor classification.

Breast cancer remains one of the leading causes of mortality among women worldwide and represents a major global health challenge. Accurate classification of breast tumors as benign or malignant is therefore of critical importance for timely diagnosis and effective treatment. This study aims to enhance breast cancer risk classification by integrating machine learning (ML) techniques with a genetic algorithm-based feature selection method. Initially, multiple ML algorithms are applied to features extracted from digitized images obtained through fine-needle aspiration (FNA) of breast masses. Subsequently, a genetic algorithm-based feature selection approach is employed to identify a subset of the most discriminative features. The results demonstrate that ML models utilizing the feature subsets selected by the genetic algorithm consistently achieve higher classification accuracy compared to their baseline counterparts. This highlights the effectiveness of the proposed feature selection strategy in improving the discriminative capacity of ML models. Beyond the observed improvements in accuracy, the refined ML models developed in this study show potential for more precise and reliable breast cancer diagnoses. By enhancing the performance of ML-based decision support systems, the genetic algorithm-based feature selection approach may contribute to the advancement of personalized treatment strategies in breast cancer care.

Coffee is an agricultural commodity of fundamental and considerable economic importance on the global market. In this study, the coffee bean varieties were examined from images via artificial intelligence due to their quality and value on the market. This study aims to create an automated system that can efficiently identify coffee beans without requiring a significant amount of time. In this study, five pre-trained Convolutional Neural Network (CNN) architectures were performed to detect four varieties of coffee beans through images. Extracting features from images is a challenging and specialized task. However, CNN possesses the ability to extract features automatically. Therefore, these architectures were employed as both deep feature extractors and classifiers. Primarily, 1600 coffee beans' images were split into 75:25 training and testing sets. Next, 5-fold cross-validation was applied during the training process. This study presented both validation and testing results. Eventually, ShuffleNet achieved the best classification performance with 99.33% and 99.75% accuracy rates in identifying types of coffee beans for the training and testing sets, respectively. As a result, this study has demonstrated that deep learning technologies can automatically recognize the different types of coffee beans.