A Novel Fusion of Computer Vision Capabilities and Deep Learning Techniques for OCT Retinal Disease Diagnosis
Pertanika Journal of Science and Technology · 30.10.2026 · 34 (5)
DOI: 10.47836/pjst.34.5.19 ↗Kurz gefasst
Fachartikel in der Zeitschrift Pertanika Journal of Science and Technology, online erschienen am 30.10.2026. Verfasst von Nagham Salim Mohammed und 2 weiteren Autor:innen.
Abstract (Originalsprache)
The main objective is to address the classification limitations of traditional models for diagnosing retinal macular diseases, including neovascularisation (CNV), diabetic macular oedema (DME), and age-related macular degeneration (AMD) associated with drusen. An Optical Coherence Tomography (OCT)- based computer vision framework has been proposed to classify these diseases. The models achieved this by addressing the inability of traditional models to learn spatial dependencies and the need for an inductive bias to learn fine retinal texture patterns in Vision Transformer (ViT) models. A dual-model HR50_ViTB/16 was proposed, comprising a full-depth ViTB/16 model for capturing spatial information, a pre-trained ResNet50 model for extracting local, multiscale hierarchical information to further improve performance, and a learnable merging layer to integrate local and global information extracted from both branches. The proposed framework outperformed all traditional models, achieving an overall accuracy of 96.14% and an F1 score of 0.952 for the DRUSEN class, which is the most difficult and has very limited data for classification. Its significance lies in the fact that these results demonstrate the quality of the local and global capabilities of ResNet50 networks and ViTB/16 transformers, respectively, which can complement each other to overcome the limitations of each model.
Zitieren
APA
Mohammed, N. S., Almolaa, O. S., & Alsaif, O. I. (2026). A Novel Fusion of Computer Vision Capabilities and Deep Learning Techniques for OCT Retinal Disease Diagnosis. Pertanika Journal of Science and Technology, 34(5). https://doi.org/10.47836/pjst.34.5.19
Vancouver
Mohammed NS, Almolaa OS, Alsaif OI. A Novel Fusion of Computer Vision Capabilities and Deep Learning Techniques for OCT Retinal Disease Diagnosis. Pertanika Journal of Science and Technology. 2026;34(5). doi:10.47836/pjst.34.5.19
BibTeX
@article{Mohammed2026,
author = {Mohammed, Nagham Salim and Almolaa, Omar S. and Alsaif, Omar Ibrahim},
title = {A Novel Fusion of Computer Vision Capabilities and Deep Learning Techniques for OCT Retinal Disease Diagnosis},
journal = {Pertanika Journal of Science and Technology},
year = {2026},
volume = {34},
number = {5},
doi = {10.47836/pjst.34.5.19}
}
EUVIMED