ABRAR ALAIDAROS, AREEJ ALSHUTAYRI, REHAB QAROUT, Aisha Blfgeh, Ahmed Alamri and Enas Jambi
Adv. Artif. Intell. Mach. Learn., 5 (4):4433-4443
1. ABRAR ALAIDAROS: University of Jeddah, College of Computer Science and Engineering, Jeddah, Saudi Arabia.
2. AREEJ ALSHUTAYRI: Department of Computer Science and Artificial Intelligence, College of Computer Science and Engineering, University of Jeddah, Jeddah, Saudi Arabia
3. REHAB QAROUT: Department of Computer Science and Artificial Intelligence, College of Computer Science and Engineering, University of Jeddah, Jeddah, Saudi Arabia
4. Aisha Blfgeh: Department of Computer Science and Artificial Intelligence, College of Computer Science and Engineering, University of Jeddah, Jeddah, Saudi Arabia
5. Ahmed Alamri: Department of Information System and Technology, College of Computer Science and Engineering, University of Jeddah Jeddah, Saudi Arabia
6. Enas Jambi: Department of Computer Science and Artificial Intelligence, College of Computer Science and Engineering, University of Jeddah Jeddah, Saudi Arabia
DOI: 10.54364/AAIML.2025.54246
Article History: Received on: 03-Aug-25, Accepted on: 22-Oct-25, Published on: 29-Oct-25
Corresponding Author: ABRAR ALAIDAROS
Email: AALAIDAROS0003.stu@uj.edu.sa
Citation: Abrar Alaidaros, et al. Diabetic Retinopathy Detection Using a Hybrid Model. Advances in Artificial Intelligence and Machine Learning. 2025;5(4):246. https://dx.doi.org/10.54364/AAIML.2025.54246
Among people with diabetes, diabetic retinopathy is considered the main cause of vision loss. Early detection is crucial to prevent vision loss. To identify diabetic retinopathy, A regular scheduled checkup is critical before the occurrence of vision loss. Because it progresses through multiple stages and the early stages usually do not contain noticeable symptoms, routine eye examinations are required. Relying on manual checkups can be prone to human error, and it is time-consuming. This can postpone the initiation of necessary interventions. Automated Artificial Intelligence (AI) based detection is therefore essential. This study aimed to explore further the Convolutional Neural Network (CNN) and the Singular Value Decomposition (SVD), Support Vector Machine(SVM) model, combining CNN for feature extraction, SVD for reducing the number of features, and SVM for classification. The model was applied to the Messidor dataset, which consists of images of the retina. The model achieved 69% accuracy.