Deep learning applications in OCT and OCT angiography for retinal disease diagnosis: A review

Audrey, Huong and Muhammad-Sukki1, Jalaluddin and Loh, Ser Lee and Muhammad-Sukki, Firdaus and Xavier, Ngu (2026) Deep learning applications in OCT and OCT angiography for retinal disease diagnosis: A review. Applications Of Modelling And Simulation, 10. pp. 196-212. ISSN 2600-8084

Text 0234005082026824263499.pdf Available under License Creative Commons Attribution. Download (1MB)

Abstract

Retinal diseases are among the most pressing global health challenges for which novel diagnostic approaches are urgently needed. Optical Coherence Tomography (OCT) and OCT Angiography (OCTA) are novel imaging techniques that provide high-resolution visualization of retinal microvascular abnormalities. Nonetheless, interpreting complex OCT/OCTA images requires extensive knowledge and experience, prompting the need for automated analysis using deep learning technology. This review provides a critical overview of recent progress in deep learning applications for various OCT and OCTA data analysis, including the model training paradigms, optimization techniques, and image processing applications. Three hundred seventy-three articles were found based on the keywords “retinal disease”, “OCT” or “OCTA”, and “deep learning” in literature searches. Following title, abstract, and keyword screening, 170 articles were reviewed in full text for eligibility, with 27 studies included in the final review. Selected articles were reviewed to determine the dataset used, the employed deep learning and image preprocessing methods, and the evaluated performance metrics. This paper summarizes the applications of deep learning in improving the performance of retinal disease classification and identifies future directions for enhancing the reliability of AI-based models. Finally, this review outlines the challenges associated with current research andexisting strategies, and suggests future directions to enhance the medical relevance of deep-learning diagnostic systems.

Item Type:	Article
Uncontrolled Keywords:	Classification, Deep learning, Optical coherence tomography, Retinal disease.
Divisions:	Faculty Of Electrical Technology And Engineering
Depositing User:	Norfaradilla Idayu Ab. Ghafar
Date Deposited:	04 Sep 2026 01:07
Last Modified:	04 Sep 2026 01:07
URI:	http://eprints.utem.edu.my/id/eprint/30374
Statistic Details:	View Download Statistic

Actions (login required)

View Item