Yazar "Cetintas, Dilber" seçeneğine göre listele

Listeleniyor 1 - 5 / 5
  • Küçük Resim Yok
    Öğe
    A Deep Learning Approach to Alzheimer's Diagnosis Using EEG Data: Dual-Attention and Optuna-Optimized SVM
    (Mdpi, 2025) Arikan, Funda Bulut; Cetintas, Dilber; Aksoy, Aziz; Yildirim, Muhammed
    Background/Objectives: Alzheimer's disease (AD) is a progressive neurodegenerative disorder, pathologically defined by the accumulation of amyloid-beta plaques and tau-related neurofibrillary tangles in the brain. It represents a principal driver of cognitive deterioration in middle-aged and elderly populations. Early diagnosis and pharmacological management of the disease markedly improve both the quality and duration of life. Methods: Electroencephalography (EEG) is critical in detecting and analyzing Alzheimer's disease. The widespread use of mobile EEG devices in recent years has necessitated real-time and effective data processing. However, extracting disease-specific features from EEG data still poses a significant challenge, especially in cases that must be completed quickly. This study aims to determine the frequency bands associated with Alzheimer's disease in EEG data obtained from multiple channels and to accelerate the detection methods. An accurate classification that requires little computation is the primary goal. Results: EEG recordings of 48 individuals (24 AD and 24 healthy controls (HC)) obtained from Florida State University were divided into Alpha, Beta, Delta, Gamma, and Theta frequency bands; scalograms and spectrograms were generated for each frequency band. The effectiveness of these bands was evaluated using the MobileNetV2 architecture. The results showed that Delta and Beta frequency bands were the most significant for Alzheimer's detection. By analyzing the features obtained from the Delta and Beta bands using the MobileNetV2 model integrated with the Dual-Attention Mechanism, it was determined that the attention mechanisms improved model performance by 2%. In addition, the use of an SVM classifier with hyperparameters optimized via Optuna resulted in approximately 3% performance improvement, suggesting that hyperparameter tuning may contribute positively to classification accuracy. Furthermore, combining features obtained from these frequency bands increased the detection performance when evaluated with larger datasets. Conclusions: The study demonstrates the potential of frequency band-based analyses and feature fusion methods to increase the accuracy and efficiency of Alzheimer's diagnosis using EEG data. The results are promising; however, they should be interpreted with caution regarding their generalizability.
  • Küçük Resim Yok
    Öğe
    Dense-MoE vs Lite-MoE: A Gating-Weight-Aware Pruning Framework for Unpaired Multimodal Breast Cancer Diagnosis
    (Springer, 2026) Cetintas, Dilber; Tuncer, Taner; Kilicarslan, Gulhan; Sevi, Mehmet
    This study proposes a unique Mixture-of-Experts (MoE)-based deep learning framework for the effective use of unpaired multimodal images in breast cancer diagnosis. Mammography (MG), ultrasonography (US), and magnetic resonance imaging (MRI) data are modeled as independent expert networks based on DenseNet-121, and a Gating Network mechanism is integrated to dynamically determine the contribution of each modality in the decision-making process. Thus, the model offers a flexible and clinically relevant structure without requiring all modalities to be available in every patient. The unique contribution of the study, the Gating-Weight-Aware Selection strategy, performs pruning by considering not only the individual Area Under the Curve (AUC) performance of the experts but also their usage rate within the model. Experimental results show that the Dense-MoE model achieves an AUC of 0.9661 and 89% accuracy. The pruned Lite-MoE model, despite reducing the number of parameters by approximately 33%, largely maintains its performance with an AUC of 0.9529 and 88% accuracy. Clinical evaluations of the proposed model were examined using Gradient-weighted Class Activation Mapping (GradCAM) heatmaps. The results showed that the model has the potential to be a flexible, scalable, and high-performance clinical decision support system in scenarios where data is incomplete.
  • Küçük Resim Yok
    Öğe
    Detection of Diffusion-Generated Images Using CBAM-Enhanced Pre-Trained CNNs
    (Istanbul Univ, 2026) Cetintas, Dilber; Yucel, Zehra
    AI-generated images have amplified the need for effective methods to distinguish between real and synthetic visuals. This underscores the need to develop new approaches to ensure data integrity and combat misinformation. While the existing literature predominantly focuses on Generative Adversarial Networks (GAN)-based synthetic images, researchers have largely overlooked the detection of diffusionbased models. This study fills this gap by demonstrating the potential of convolutional block attention module (CBAM)-enhanced convolutional neural networks (CNNs) for the effective detection of diffusionbased synthetic images. In this study, we use CNNs enhanced with the CBAM to propose a novel approach for detecting synthetic images. The CBAM-enhanced model, trained on the CIFAKE dataset, achieved a remarkable accuracy of97.38% in detecting synthetic images. We integrate pre-trained CNN architectures, such as ResNet50 and DenseNet121, with a CBAM attention mechanism, which enhances performance by focusing on salient spatial and channel information. This approach presents a model that significantly enhances the detection capabilities for distinguishing fake images. Ourfindings contribute to the field of deepfake detection by providing a robust solution for automated digital image vetting, with implications for AI ethics, security, and broader societal discourse. The implementation details and source code are available at https://github.com/cmpe-dev/Fake-Detector-with-CBAM.
  • Küçük Resim Yok
    Öğe
    Fusion-Based Deep Learning Approach for Renal Cell Carcinoma Subtype Detection Using Multi-Phasic MRI Data
    (Mdpi, 2025) Kilicarslan, Gulhan; Cetintas, Dilber; Tuncer, Taner; Yildirim, Muhammed
    Background/Objectives: Renal cell carcinoma (RCC) is a malignant disease that requires rapid and reliable diagnosis to determine the correct treatment protocol and to manage the disease effectively. However, the fact that the textural and morphological features obtained from medical images do not differ even among different tumor types poses a significant diagnostic challenge for radiologists. In addition, the subjective nature of visual assessments made by experts and interobserver variability may cause uncertainties in the diagnostic process. Methods: In this study, a deep learning-based hybrid model using multiphase magnetic resonance imaging (MRI) data is proposed to provide accurate classification of RCC subtypes and to provide a decision support mechanism to radiologists. The proposed model performs a more comprehensive analysis by combining the T2 phase obtained before the administration of contrast material with the arterial (A) and venous (V) phases recorded after the injection of contrast material. Results: The model performs RCC subtype classification at the end of a five-step process. These are regions of interest (ROI), preprocessing, augmentation, feature extraction, and classification. A total of 1275 MRI images from different phases were classified with SVM, and 90% accuracy was achieved. Conclusions: The findings reveal that the integration of multiphase MRI data and deep learning-based models can provide a significant improvement in RCC subtype classification and contribute to clinical decision support processes.
  • Küçük Resim Yok
    Öğe
    Vocal Fold Disorders Classification and Optimization of a Custom Video Laryngoscopy Dataset Through Structural Similarity Index and a Deep Learning-Based Approach
    (Mdpi, 2025) Emre, Elif; Cetintas, Dilber; Yildirim, Muhammed; Emre, Sadettin
    Background/Objectives: Video laryngoscopy is one of the primary methods used by otolaryngologists for detecting and classifying laryngeal lesions. However, the diagnostic process of these images largely relies on clinicians' visual inspection, which can lead to overlooked small structural changes, delayed diagnosis, and interpretation errors. Methods: AI-based approaches are becoming increasingly critical for accelerating early-stage diagnosis and improving reliability. This study proposes a hybrid Convolutional Neural Network (CNN) architecture that eliminates repetitive and clinically insignificant frames from videos, utilizing only meaningful key frames. Video data from healthy individuals, patients with vocal fold nodules, and those with vocal fold polyps were summarized using three different threshold values with the Structural Similarity Index Measure (SSIM). Results: The resulting key frames were then classified using a hybrid CNN. Experimental findings demonstrate that selecting an appropriate threshold can significantly reduce the model's memory usage and processing load while maintaining accuracy. In particular, a threshold value of 0.90 provided richer information content thanks to the selection of a wider variety of frames, resulting in the highest success rate. Fine-tuning the last 20 layers of the MobileNetV2 and Xception backbones, combined with the fusion of extracted features, yielded an overall classification accuracy of 98%. Conclusions: The proposed approach provides a mechanism that eliminates unnecessary data and prioritizes only critical information in video-based diagnostic processes, thus helping physicians accelerate diagnostic decisions and reduce memory requirements.