Emir Oncu

Mr. Emir Oncu

Bioengineering, Yıldız Technical University, Istanbul, İstanbul, Turkey (Türkiye)
orcid
Biomedical Engineer and Researcher

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications

29 April 2025 in Computers in Biology and Medicine
Verified
Journal Article
Closed access
Emir OncuEmir Oncu
Emir Oncu
KAKadriye Yasemin Usta Ayanoglu
Kadriye Yasemin Usta Ayanoglu
Fatih CiftciFatih Ciftci
Fatih Ciftci

20 May 2025 in Bone Reports
Verified
Journal Article
Open access
SBSema Ertan Birsel
Sema Ertan Birsel
EDEkrem Demirci
Ekrem Demirci
ASAli Seker
Ali Seker
KAKadriye Yasemin Usta Ayanoğlu
Kadriye Yasemin Usta Ayanoğlu
Emir OncuEmir Oncu
Emir Oncu
Fatih CiftciFatih Ciftci
Fatih Ciftci
Hip displacement is a significant concern in children with cerebral palsy (CP), necessitating accurate and timely assessment to prevent long-term complications. This study developed a support vector machine (SVM) model to classify hip conditions using migration percentage (MP) measurements obtained from 176 hips across 88 anteroposterior pelvic radiographs. MP values were categorized into three groups: normal (MP ≤ 30 %), risky (30 % < MP ≤ 60 %), and dislocated (MP > 60 %). The SVM model was evaluated using stratified k-fold cross-validation, with accuracy, precision, recall, and F1-scores as key metrics. Its classifications were compared to manual evaluations performed by an orthopedic resident and a pediatric orthopedic surgeon. The model achieved an overall accuracy of 92.898 %, surpassing the consistency and reliability of manual assessments, particularly in identifying dislocated hips. Statistical analysis showed no significant differences between the model's MP measurements and those of the clinicians, validating its effectiveness. This study highlights the potential of SVM models to enhance diagnostic accuracy, reduce variability in evaluations, and support clinical decision-making. Future research should expand the dataset and incorporate advanced machine learning models to further improve diagnostic precision.

3 April 2025 in Applied Sciences
Verified
Journal Article
Open access
PDF file
AYAslıhan Yelkenci
Aslıhan Yelkenci
GPGünseli Güven Polat
Günseli Güven Polat
Emir OncuEmir Oncu
Emir Oncu
Fatih CiftciFatih Ciftci
Fatih Ciftci
Space maintainers (SMs) are essential for preserving dental arch integrity after premature tooth loss. This study aimed to develop a deep learning model to predict the necessity of SMs and identify specific teeth requiring intervention. A dataset of 400 dental X-rays was preprocessed to standardize image dimensions and convert them into numerical representations for machine learning. The dataset was divided into training (80%) and testing (20%) subsets. A Convolutional Neural Network (CNN) was designed with multiple convolutional and pooling layers, followed by fully connected layers for binary classification. The model was trained using 30 epochs and evaluated with accuracy, precision, recall, F1-score, ROC AUC, and MCC. The CNN achieved 94% accuracy, with a precision of 0.93 for Class 0 (no SM needed) and 0.95 for Class 1 (SM needed). The ROC AUC was 0.94, and the MCC was 0.875, indicating strong reliability. When tested on 86 X-ray images, the model successfully identified specific teeth (showing teeth number) requiring SMs, with minimal errors. These results suggest that the proposed AI model provides high-performance predictions for SM necessity, offering a valuable decision-support tool for pediatric dentistry.

2 January 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
Emir OncuEmir Oncu
Emir Oncu
Background and Objective: Alzheimer’s disease (AD) is a challenging neurodegenerative disorder to diagnose, necessitating innovative solutions for early detection and classification. Traditional diagnostic methods often lack sensitivity or scalability, highlighting the need for advanced approaches. This study proposes a dual-model framework integrating an Artificial Neural Network (ANN) and a Convolutional Neural Network (CNN) to enhance diagnostic accuracy. Methods: The framework combines two AI models. The ANN was trained on clinical data from 1,200 patients, incorporating 31 demographic, symptomatic, and behavioral features, to assess Alzheimer’s risk. The CNN analyzed 4,876 Magnetic Resonance Imaging (MRI) images to confirm the diagnosis and classify the disease into four stages: mild demented, moderate demented, very mild demented, and non-demented. Grad-CAM visualizations enhanced interpretability, offering clinically relevant insights. Results: The ANN model achieved an accuracy of 87.08% in assessing Alzheimer’s risk, while the CNN model excelled with a 97% accuracy in disease staging. Grad-CAM visualizations highlighted critical regions in the MRI images, enhancing the transparency and reliability of the diagnostic process. The results demonstrate the complementary strengths of both models in providing a comprehensive diagnostic solution. Conclusion: The integrated ANN-CNN framework shows promise in revolutionizing Alzheimer’s diagnostics by combining clinical and imaging data for accurate detection. While limited by MRI availability and variability in clinical data, the framework underscores AI's potential in advancing neurodegenerative disease diagnosis. Future directions include integrating wearable technology and lightweight CNNs to improve scalability, accessibility, and early intervention.

2 December 2024 in Preprints.org
Verified
Preprint
Open access
PDF file
Emir OncuEmir Oncu
Emir Oncu
Lung cancer remains a leading cause of cancer-related mortality worldwide, underscoring the critical need for early and accurate diagnostic tools. This study presents a novel multimodal artificial intelligence (AI) framework integrating convolutional neural networks (CNNs) and artificial neural networks (ANNs) to enhance lung cancer detection and classification. The CNN model was trained on a dataset of 900 CT images, meticulously preprocessed and categorized into four classes: adenocarcinoma, large cell carcinoma, squamous cell carcinoma, and normal tissue. Using advanced techniques such as Grad-CAM (Gradient-weighted Class Activation Mapping), the CNN provided interpretable predictions with a weighted average accuracy of 91%, achieving near-perfect AUC values for most classes. Simultaneously, an ANN was developed using clinical data from 999 patients, incorporating 24 features that captured demographic, symptomatic, lifestyle, and genetic factors. The ANN exhibited an exceptional classification performance, attaining an overall accuracy of 99% with minimal misclassifications. Together, these models form a robust diagnostic system that leverages the complementary strengths of imaging and clinical data analysis to improve diagnostic accuracy and clinical decision-making. The proposed framework not only delivers precise classification but also emphasizes interpretability, scalability, and clinical applicability. While the CNN excels in imaging-based diagnosis, the ANN offers reliable predictions grounded in patient-specific clinical data. Future work will focus on expanding datasets, incorporating additional diagnostic modalities, and optimizing the models for real-world clinical deployment. This study highlights the transformative potential of multimodal AI in advancing the early detection and management of lung cancer.