Sürdürülebilir Kalkınma Amaçları

• 6
• 3
• 2
• 2
• 2
• 1

• MakaleAçık Erişim2026Scopus SJRQ2/r/munzur/1262
  A Genetic Algorithm-Optimized ConvNeXtV2-YOLOv8 Framework for Dental Caries Detection in Panoramic Radiographs

  Sobahi, Nebras; Küçük, Deniz Bora; Kılıç, Kazım; İmak, Andaç; Çelebi, Adalet; Alghamedi, Yazyd; Yazicioglu, Cafer; Türkoğlu, Muammer; Şengür, Abdulkadir

  Background/Objectives: Dental caries is one of the most common oral diseases worldwide, and early diagnosis is essential for effective treatment. However, detecting carious lesions in panoramic radiographs is challenging because of low image contrast, anatomical complexity, and overlapping structures. This study aimed to develop an improved object detection framework for dental caries localization in panoramic radiographs. Methods: A modified YOLOv8 architecture was developed by integrating ConvNeXtV2 blocks into the Cross-Stage Partial (C2f) modules to enhance feature extraction and gradient propagation. To improve detection performance and reduce overfitting, key training hyperparameters were optimized using a genetic algorithm. The proposed model was evaluated on a dataset of 474 panoramic dental radiographs annotated by experts using bounding boxes. Results: The optimized model achieved a box precision of 78.4%, a recall of 53.6%, and a mean average precision at 50% intersection-over-union threshold (AP50) of 62.6%. Compared with the baseline YOLOv8 model, the proposed approach improved precision and AP50. Visual and quantitative analyses demonstrated that the ConvNeXtV2-enhanced architecture enabled more accurate localization of carious regions. Conclusions: The results indicate that combining ConvNeXtV2-based architectural enhancement with genetic algorithm-based hyperparameter optimization is an effective strategy for dental caries localization in panoramic radiographs. Although recall remains a limitation, the proposed framework shows potential as a computer-aided diagnostic tool for supporting clinical caries assessment.

  • 
  İncele
• MakaleYalnız künye2026Scopus SJRQ1/r/munzur/1174
  Modeling stack voltage and remaining useful life of proton exchange membrane fuel cells using deep neural networks

  Dogan, Gurkan; Imak, Andaç

  Currently, various energy conversion systems are being developed to reduce environmental impacts. Proton exchange membrane fuel cells (PEMFCs), which stand out with their high efficiency characteristics, have emerged as an important alternative. However, these advantages, along with high costs and uncertainties regarding remaining useful life (RUL) estimation, restrict their further commercialization, especially in the automotive industry. This study presents a forecasting framework using a long short-term memory (LSTM) architecture enhanced with the Bahdanau attention mechanism for predicting RUL and long-term degradation trends. The model thus exhibits a performance that can more effectively capture dependencies in time series and adaptively weight significant time steps. Furthermore, the tree-structured parzen estimator (TPE) was chosen for the appropriate selection of hyperparameters in the developed forecasting framework. The validity, stability, and performance of the proposed model under different operating conditions were examined on the FC1 and FC2 datasets. Experimental studies have shown that the proposed method is the most successful among the best methods, achieving RE metric values of 1.628 × 10−3, 0.003 × 10−1, 99.16 × 10−2, and 0.044 × 10−1 respectively in estimating the best RMSE, MAPE, R2, and RUL.

  • 
  • 
  İncele
• MakaleYayıncıda açık erişim2026Scopus SJRQ1/r/munzur/1123
  Explainable AI-based deep learning architectures for hydrogen prediction in biomass syngas and photocatalysis processes: Towards a sustainable and efficient energy economy

  Imak, Andaç

  This study, focusing on sustainable energy production and particularly the hydrogen (H2) economy, presents an advanced artificial intelligence (AI) framework for accurately predicting H2 production from sucrose photocatalysis catalyzed by LaFeO3 (GLFO) and the H2 concentration in syngas derived from biomass conversion. The research was conducted along two main directions: (i) deep learning (DL) architectures comprising nine specially designed recurrent neural network-based models and (ii) six conventional machine learning (ML) algorithms. For both H2 production systems, the best-performing models were identified and further interpreted using explainable AI (XAI) based on SHapley Additive exPlanations (SHAP). Among all investigated algorithms, a hybrid One-dimensional Convolutional Neural Network–Bidirectional Long Short-Term Memory–Bidirectional Gated Recurrent Unit (1D-CNN–BiLSTM–BiGRU) model demonstrated the highest predictive performance. Using 10-fold cross-validation, the proposed model achieved outstanding accuracy, with an R2 value of 0.9987 ± 0.0011 for photocatalytic H2 production prediction and 0.9980 ± 0.0006 for syngas H2 concentration prediction. These results highlight the potential of the proposed framework as a precise, reliable, and practical methodological approach for modeling and optimizing complex thermochemical energy systems.

  • 
  • 
  • 
  İncele
• MakaleAçık Erişim2025Scopus SJRQ2/r/munzur/1479
  Hybrid CNN-Transformer Model for Accurate Impacted Tooth Detection in Panoramic Radiographs

  Küçük, Deniz Bora; Imak, Andaç; Özçelik, Salih Taha Alperen; Çelebi, Adalet; Türkoğlu, Muammer; Sengur, Abdulkadir; Koundal, Deepika

  Background/Objectives: The integration of digital imaging technologies in dentistry has revolutionized diagnostic and treatment practices, with panoramic radiographs playing a crucial role in detecting impacted teeth. Manual interpretation of these images is time consuming and error prone, highlighting the need for automated, accurate solutions. This study proposes an artificial intelligence (AI)-based model for detecting impacted teeth in panoramic radiographs, aiming to enhance accuracy and reliability. Methods: The proposed model combines YOLO (You Only Look Once) and RT-DETR (Real-Time Detection Transformer) models to leverage their strengths in real-time object detection and learning long-range dependencies, respectively. The integration is further optimized with the Weighted Boxes Fusion (WBF) algorithm, where WBF parameters are tuned using Bayesian optimization. A dataset of 407 labeled panoramic radiographs was used to evaluate the model’s performance. Results: The model achieved a mean average precision (mAP) of 98.3% and an F1 score of 96%, significantly outperforming individual models and other combinations. The results were expressed through key performance metrics, such as mAP and F1 scores, which highlight the model’s balance between precision and recall. Visual and numerical analyses demonstrated superior performance, with enhanced sensitivity and minimized false positive rates. Conclusions: This study presents a scalable and reliable AI-based solution for detecting impacted teeth in panoramic radiographs, offering substantial improvements in diagnostic accuracy and efficiency. The proposed model has potential for widespread application in clinical dentistry, reducing manual workload and error rates. Future research will focus on expanding the dataset and further refining the model’s generalizability.

  • 
  AtıfScopus36·OpenAlex34

  İncele
• MakaleAçık Erişim2025Scopus SJRQ2/r/munzur/1200
  HASK-Net: new hybrid attention selective kernel network for automatic colon cancer detection from colonoscopy images

  Imak, Andaç; Doğan, Gürkan; Sengur, Abdulkadir; Ergen, Burhan

  Colorectal cancer is a frequently analyzed cancer type seen today. It is the third leading of cancer in terms of incidence. It is among the second leading cases resulting in death. However, when detecting pre-cancerous polyps using traditional methods, the rate of being overlooked is high. Therefore, the rate of missed polyps can be greatly reduced with deep learning-based computer-aided diagnosis (CADx) methods. In this study, a novel convolutional neural network based on CADx, called HASK-Net, is proposed for automatic polyp detection. While HASK-Net uses ResNet50 as a feature extractor in the encoder network, the novely designed hybrid attention selective kernel convolution uses it in the decoder network to increase the representation power in feature maps and learn more complex features. In experimental results on the publicly available Hiper-Kvasir and EndoTech 2020 benchmark datasets, HASK-Net showed extremely promising performance with a dice similarity score of 97.02% and 84.83% and an mIoU score of 95.38% and 84.32%, respectively.

  • 
  İncele
• MakaleAçık Erişim2025Scopus SJRQ1/r/munzur/1090
  MobileNet- and attention-based MLP-Mixer architecture for geographical-region recognition of a marine fish (Perciformes: Carangidae) using otolith images

  Dürrani, Ömerhan; İşgüzar, Seda; İmak, Andaç; Ateşşahin, Tuncay; Cömert, Zafer; Dürrani, Syeda Zahra; Türkoğlu, Muammer

  Fishery management is crucial to sustain marine ecosystems by preventing overfishing and ensuring a fair distribution of fishing quotas. Accurately identifying the geographical origins of fish stocks is a key challenge in region-specific management strategies. Otoliths, calcified structures found in the heads of all fish species (except sharks and rays), provide insights into the life history and geographical origins of these fish. Traditional otolith analysis is time-consuming and error-prone because of manual inspection. Our study presents a novel approach using deep learning and computer vision to automate the geographical recognition of fish using otolith images. We propose a model that integrates MobileNet, which is known for its efficiency, with an advanced Mlp-Mixer that incorporates an attention mechanism to extract enhanced features. When tested on a diverse dataset of otolith images from five regions, the proposed model achieved a remarkable 96% accuracy, significantly outperforming traditional methods. This high accuracy demonstrates the potential to revolutionise fishery management by providing a fast, reliable, and automated solution for geographical region identification. In conclusion, the proposed method demonstrates the transformative potential of combining MobileNet and an attention-based Mlp-Mixer for automated fish geographic recognition using otolith images. This innovative method addresses the limitations of traditional manual inspection and paves the way for more effective and sustainable fishery management practices.

  • 
  AtıfScopus4·OpenAlex4

  İncele
• MakaleAçık Erişim2024/r/munzur/6301
  Automatic Classification of Defective Photovoltaic Module Cells Based on a Novel CNN-PCA-SVM Deep Hybrid Model in Electroluminescence Images

  Imak, Andac

  In today’s world, the rapid development of photovoltaic (PV) power plants has facilitated sustainable energy production. Maintenance and defect detection play crucial roles in ensuring the continuity of energy production. The manual inspection of electroluminescence (EL) images of PV modules requires significant human power and time investment. This study presents a method for the automatic fault detection of PV cells in EL images using hybrid deep features optimized with a principal component analysis (PCA) feature selection algorithm. A lightweight and high-performance model that combines the strengths of convolutional neural network (CNN) architectures was proposed. First, data augmentation techniques were employed owing to the imbalance between the defective and functional classes in the dataset containing EL images. In experimental studies conducted by integrating the PCA algorithm into MobileNetV2, DenseNet201, and InceptionV3 CNN models, accuracy, precision, recall, and F1-score values of 92.19%, 92%, 90%, and 91%, respectively, were achieved. When the results were analyzed, it was observed that the proposed method was effective in detecting faults in PV panel cells.

  • 
  • 
  AtıfOpenAlex8·TR Dizin2

  İncele
• MakaleAçık Erişim2024Scopus SJRQ1/r/munzur/2014
  A new hybrid approach for grapevine leaves recognition based on ESRGAN data augmentation and GASVM feature selection

  Doğan, Gürkan; Imak, Andaç; Ergen, Burhan; Sengur, Abdulkadir

  Grapevine leaf is a commodity that is collected only once a year and has a high return on investment due to its export. However, only certain types of grapevine leaves are consumed. Therefore, it is extremely important to distinguish the types of grapevine leaves. In particular, performing this process automatically on industrial machines will reduce human errors, workload, and thus cost. In this study, a new hybrid approach based on a convolutional neural network is proposed that can automatically distinguish the types of grapevine leaves. In the proposed approach, firstly, the overfitting of network models is prevented by applying data augmentation techniques. Second, new synthetic images were created with the ESRGAN technique to obtain detailed texture information. Third, the top blocks of the MobileNetV2 and VGG19 CNN models were replaced with the newly designed top block, effectively extracting features with the data. Fourthly, the GASVM algorithm was adapted and used to create a subset of the features to eliminate the ineffective and unimportant ones from the obtained features. Finally, SVM classification was performed with the feature subset consisting of 314 features, and approximately 2% higher accuracy and MCC score were obtained compared to the approaches in the literature.

  • 
  AtıfScopus8·WoS6·OpenAlex9

  İncele
• MakaleYalnız künye2024Scopus SJRQ2/r/munzur/1899
  Maxillary sinus detection on cone beam computed tomography images using ResNet and Swin Transformer-based UNet

  Çelebi, Adalet; Imak, Andaç; Üzen, Hüseyin; Budak, Ümit; Türkoğlu, Muammer; Hanbay, Davut; Şengür, Abdulkadir

  Objectives: This study, which uses artificial intelligence-based methods, aimed to determine the limits of pathologic conditions and infections related to the maxillary sinus in cone beam computed tomography (CBCT) images to facilitate the work of dentists. Methods: A new UNet architecture based on a state-of-the-art Swin transformer called Res-Swin-UNet was developed to detect the sinus. The encoder part of the proposed network model consists of a pre-trained ResNet architecture, and the decoder part consists of Swin transformer blocks. Swin transformers achieve powerful global context properties with self-attention mechanisms. Because the output of the Swin transformer generates sectorized features, the patch expanding layer was used in this section instead of the traditional upsampling layer. In the last layer of the decoder, sinus diagnosis was conducted through classical convolution and sigmoid function. In experimental works, we used a data set including 298 CBCT images. Results: The Res-Swin-UNet model achieved more success, with a 91.72% F1-score, 99% accuracy, and 84.71% IoU, outperforming the state-of-the-art models. Conclusions: The deep learning-based model proposed in the present study can assist dentists in automatically detecting the boundaries of pathologic conditions and infections within the maxillary sinus based on CBCT images.

  • 
  AtıfScopus21·OpenAlex20

  İncele
• MakaleAçık Erişim2023/r/munzur/6480
  A New Method Based on Extracting, Combining and Selecting Deep Features from Natural and Synthetic Data for Classification of Grapevine Leaf Species

  Sengur, Abdulkadir; Ergen, Burhan; Doğan, Gürkan; Imak, Andac

  In recent years, the production and consumption of grapevine leaves containing rich minerals, dietary fiber and vitamins has been intensively realized in Turkey. However, the demand for stuffed food made from grapevine leaves in the convenience food sector increases the export opportunities to different countries In commercial agricultural activities such as this, it is important to determine quality standards for sustainable marketing. Researchers have made positive progress with deep learning in smart farming applications. In this study, we propose a new method for recognizing the type of grapevine leaves to be used for consumption. In the proposed method, a dataset containing 500 images of 5 different grapevine leaf species, namely Ak, Ala Idris, Büzgülü, Dimnit and Nazli, was used. From these images, 3500 images were obtained with data augmentation techniques. In addition, a dataset consisting of more detailed textures was obtained by applying the ESRGAN model to the obtained images. VGG 19 deep learning model was used to extract features from these images. The features obtained from the two separate data sets were combined. In this way, a hybrid feature extraction model was created. The best 175 feature subsets were selected using the PCA algorithm. Finally, 96.14% accuracy was calculated using a Support Vector Machine (SVM) for classification of the obtained features.

  • 
  AtıfOpenAlex10·TR Dizin8

  İncele