Doktora Tezleri

Permanent URI for this collectionhttps://hdl.handle.net/20.500.12573/5800

Browse

Browsing Doktora Tezleri by Author "Aydın, Zafer"

Filter results by typing the first few letters
Now showing 1 - 6 of 6
  • Thumbnail Image
    Doctoral Thesis
    Biyomedikal Görüntülerin Segmentasyonu için Derin Öğrenme Yaklaşımları
    (2025) Güzel, Yasin; Aydın, Zafer
    Bu tez çalışması, biyomedikal görüntü segmentasyonunun tanısal süreçlerdeki kritik rolünden hareketle, derin öğrenme tabanlı modellerin potansiyelini üç farklı ve zorlu tıbbi alanda araştırmaktadır. Üç farklı klinik senaryoda yürütülen çalışmalarda, sırasıyla; intraoral kamera görüntülerinden mikrobiyal dental plak, MR görüntülerinden düşük dereceli glioma tümörler ve PET/BT görüntülerinden prostat bezi segmentasyonu gerçekleştirilmiştir. Her bir problem için derin öğrenme mimarileri ile klasik bilgisayarlı görü yöntemleri karşılaştırılmış; hiperparametre optimizasyonları yapılarak adil bir değerlendirme zemini oluşturulmuştur. Bulgular, derin öğrenme modellerinin her üç alanda da klasik yöntemlere ezici bir üstünlük sağladığını tutarlı bir şekilde göstermiştir. Özellikle, diş plağı segmentasyonunda U-Net Transformer modelinin, üç uzman diş hekiminden istatistiksel olarak anlamlı düzeyde daha başarılı sonuçlar elde etmesi, bu teknolojinin klinik potansiyelini vurgulamaktadır. Benzer şekilde, düşük dereceli glioma için UNet++ ve prostat segmentasyonu için Attention U-Net mimarileri en yüksek performansı sergilemiştir. Grad-CAM ve Saliency Map ile açıklanabilirlik teknikleri ile bu yüksek skor veren modellerin karar mekanizmalarının hedeflenen anatomik bölgelere doğru odaklandığı doğrulanmış ve modelleri, kara kutu algısının ötesine taşımıştır. Bu tez, farklı tıbbi görüntüleme modalitelerinde probleme özgü önerilen derin öğrenme çözümlerinin, hem otomasyonu sağlayarak verimliliği artırma hem de uzman performansını aşarak tanısal doğruluğu iyileştirme gücünü kanıtlamaktadır.
  • Thumbnail Image
    Doctoral Thesis
    Genetik ve Enfeksiyon Hastalıklarının Tespiti için Makine Öğrenmesi Yöntemleri
    (Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2024) Işık, Yunus Emre; Aydın, Zafer
    Completion of the whole human genome in the 2003 has led to various advances in many fields, particularly in biology, genetics, health sciences, treatment, and pharmacology. In the following years, spread of faster and cheaper sequencing technologies has enabled us to extract and analyze genetic profiles of individuals digitally. Consequently, individual-specific forecasting and personalized treatment and precision medicine-, what once seemed like science fiction, have become more and more real. In both approaches, one of the crucial steps is identifying the presence of diseases using individual-specific genetic data. This thesis aims to comprehensively and comparatively evaluate the predictive performance of machine learning methods for Behçet's disease and respiratory infections. Additionally, feature selection methods were employed to identify the genetic factors (such as SNPs and genes) associated with disease presence for both diseases. Furthermore, the usability of selected features depending on biological pathway-driven active subnetworks listed in the literature was analyzed for the prediction of Behçet's disease. For the respiratory infection prediction problem, on the other hand, the prediction performance of features calculated by single-sample gene set enrichment analysis (ssGSEA) was evaluated using different machine learning methods. As the data types used in both experiments were different (genome-wide association studies data, gene expression profiles), the performance of machine learning approaches on different data types was also observed. It is hoped that the findings of both experiments will contribute to future machine learning based disease prediction studies.
  • Thumbnail Image
    Doctoral Thesis
    Histopatoloji Görüntülerinden Bilgisayar Destekli Kanser Tespiti
    (Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2023) Taşdemir, Sena Büşra Yengeç; Yılmaz, Bülent; Aydın, Zafer
    Detecting colon adenomatous polyps early is crucial for reducing colon cancer risk. This thesis investigated various deep learning approaches for computer-aided diagnosis of colon polyps on histopathology images using deep learning. The thesis addressed key challenges in polyp classification, including differentiating adenomatous polyps from non-adenomatous tissues and multi-class classification of polyp types. Initially, a histopathology image dataset is collected and refined from Kayseri City Hospital. The first study used stain normalization algorithms and an ensemble framework for binary classification, achieving 95% accuracy on the custom dataset and 91.1% and 90% on UnitoPatho and EBHI datasets, respectively. The second study implemented a tailored version of the supervised contrastive learning model for multi-class classification, outperforming state-of-the-art deep learning models with accuracies of 87.1% on custom dataset and 70.3% on UnitoPatho dataset. The third study proposed a self-supervised contrastive learning approach for utilizing all data in cases of limited labeled images. This approach achieved better performance than transfer learning with ImageNet pre-trained models. In conclusion, this PhD thesis investigated deep learning approaches for computer-aided diagnosis of colon polyps on histopathology images, demonstrating high accuracy in binary and multi-class classification, outperforming state-of-the-art models. These findings contribute to improving colon polyp classification accuracy and efficiency, ultimately facilitating the early detection and prevention of colon cancer.
  • Thumbnail Image
    Doctoral Thesis
    Protein Yapı Tahmini için Derin Öğrenme Modellerinin Geliştirilmesi
    (Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2022) Görmez, Yasin; Aydın, Zafer
    The three-dimensional structure of a protein provides important clues about the function of that protein. Although there have been many studies on protein structure prediction, the problem has still not been solved completely. As it is very difficult to predict the three-dimensional structure of a protein directly, predictions of structural properties of proteins such as secondary structure, solvent accessibility, and torsion angles are carried out first, which are later used as inputs to more elaborate structure estimation tasks. In this thesis, novel deep learning models have been developed by using convolutional neural networks (CNN), graph convolutional networks (GCN) and long-short-term memory (LSTM) recurrent neural networks to predict secondary structure, solvent accessibility and torsion angles of proteins. A rich feature set formed by using PSI-BLAST, HHBlits, physicochemical properties, structural profile matrices, AA index values, and graphs representing the relationship between amino acids were used as inputs to the models. In the first study, a deep learning model was developed by using CNN and GCN layers for secondary structure prediction. In the second study, LSTM layers were added to the first model, which was extended to make solvent accessibility and torsion angle predictions as well using the multi-task learning approach. In both studies, graphs were generated using neighborhood relations between amino acids. In the last study, a novel U-net-based model was designed for secondary structure prediction using CNN, GCN, and LSTM layers. The graph matrices used as input to GCN layers were obtained by using protein contact map prediction. All models were trained, optimized and tested on benchmark data sets. Improvements were obtained in accuracy as compared to the state-of-the-art
  • Thumbnail Image
    Doctoral Thesis
    Trafik Yoğunluğu Tahmini için Derin Öğrenme Modelleri
    (Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2024) Çini, Nevin; Aydın, Zafer
    In the last 50 years, with the growth of cities and increase in the number of vehicles and mobility, traffic has become troublesome. As a result, traffic flow prediction started to attract attention as an important research area. However, despite the extensive literature, traffic flow prediction still remains as an open research problem, specifically for long- term traffic flow prediction. Compared to the models developed for short-term traffic flow prediction, the number of models developed for long-term traffic flow prediction is very few. Based on this shortcoming, in this study, we focus on long-term traffic flow prediction and propose a novel deep ensemble model (DEM). In order to build this ensemble model, first, we developed a convolutional neural network (CNN), a long short term memory (LSTM) network, and a gated recurrent unit (GRU) network as deep learning models, which formed the base learners. In the next step, we combine the output of these models according to their individual forecasting success. We use another deep learning model to determine the success of the individual models. Our proposed model is a flexible ensemble prediction model that can be updated based on traffic data. To evaluate the performance of the proposed model, we use a publicly available dataset. Numerical results show that our proposed model performs better than individual deep learning models (i.e., LSTM, CNN, GRU), selected traditional machine learning models (i.e., linear regression (LR), decision tree regression (DTR), k-nearest-neighbors regression (KNNR) and other ensemble models such as random-forest-regression(RFR).
  • Thumbnail Image
    Doctoral Thesis
    Zamansal Bilgiden Faydalanarak Videodan Orman Yangınlarının Erken Tespiti
    (Abdullah Gül Üniversitesi, Fen Bilimleri Enstitüsü, 2022) Taş, Merve; Taşdemir, Kasım; Aydın, Zafer
    Forest fires are considered as the major threats to lives, properties and to the integrity of the ecosystem around the world. In most cases, the fire damage can be reduced, when the initial signs of the fire are detected in a timely manner. Since smoke is considered as the first visual sign of fire, detection of smoke is vital. Hence, a successfully designed smoke detection system is essentially critical in the early detection of smoke for outdoor environments. The existing smoke detection methods suffer from high false alarm rates and cannot accurately detect smoke in hazy environments. To address these problems, this thesis is focused on smoke detection model at an early stage that utilizes deep learning (DL) based techniques for outdoor locations. This work contributes mainly to four aspects of smoke detection: (1) new datasets preparation for three smoke detection tasks classification, detection-segmentation, and video classification, (2) utilizing transfer learning to detect the smoke on the relatively small dataset, (3) image dehazing process that includes removing the haze from the dataset images to enhance the system performance, (4) designing a novel hybrid video classification model by combining the two DL based video classification structures. This work will be a resourceful reference for researchers working in the fields of forest fire or smoke detection studies at an early stage. The experiments, research findings, and enhanced performance of the smoke detection system provide a source of information about smoke detection. Current studies can be utilized to further improve the design of efficient and reliable fire safety models. Keywords: Deep Learning, Spatio-Temporal Information, Forest Fire Early Detection, Smoke Detection, Image Dehazing