The utility of artificial intelligence in characterization and detecting causes of macular edema: A spectral-domain OCT-based algorithm study

• Published in: Experimental eye research Vol. 260; p. 110619
• Main Authors: Alzu'bi, Amal, Momany, Sondos, Aleshawi, Abdelwahab, Tashtoush, Mais, Al-Dwairi, Rami
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.11.2025
• Subjects: Algorithms; Artificial Intelligence; Deep Learning; Diabetic Retinopathy - complications; Diabetic Retinopathy - diagnosis; Humans; Macular Degeneration - complications; Macular Degeneration - diagnosis; Macular edema; Macular Edema - diagnosis; Macular Edema - diagnostic imaging; Macular Edema - etiology; Neural Networks, Computer; OCT; ResNet152; Retrospective Studies; Tomography, Optical Coherence - methods; OCT; Macular edema; ResNet152; Artificial intelligence
• ISSN: 0014-4835; 1096-0007; 1096-0007
• DOI: 10.1016/j.exer.2025.110619

Macular Edema (ME), a prevalent cause of vision loss, can arise from various retinal conditions, most notably diabetic macular edema (DME) and age-related macular degeneration (AMD). Accurate and timely differentiation among these causes is necessary for appropriate treatment; however, it remains a diagnostic challenge. This research addresses the gap in automated ME classification by developing and evaluating a deep learning framework capable of distinguishing between DME, AMD, and normal retinal conditions using optical coherence tomography (OCT) images. A retrospective dataset comprising 1040 OCT images from King Abdullah University Hospital (KAUH) was used in conjunction with a public dataset for benchmarking. The dataset was divided into annotated and non-annotated images, with preprocessing, augmentation, and simulated segmentation applied to improve the model performance. We benchmarked and evaluated three pretrained convolutional neural networks—ResNet152, InceptionV3, and MobileNetV2. Among the models, InceptionV3 and ResNet152 achieved the highest accuracies (95 %–98 %) across both datasets. MobileNetV2, on the other hand, showed moderate accuracy on the KAUH dataset (89 %) but exhibited strong performance on the public dataset (97 %). Explainable AI (XAI) techniques, specifically Grad-CAM, were applied to visualize the model predictions, and the outcomes were manually validated against annotated data to assess interpretability. The findings support the integration of a robust CNN architecture and XAI techniques to enhance diagnostic precision and aid clinical decision-making in ophthalmology. •The most prevalent causes of ME include DME, AMD, and retinal vein occlusion.•Deep learning has been adopted to enhance the automated diagnosis of ME using OCT scans.•Three pretrained CNN architectures—ResNet152, InceptionV3, and MobileNetV2 were evaluated.•InceptionV3 and ResNet152 achieved the highest accuracies (95 %–98 %).•XAI when combined with robust CNN models can enable accurate diagnosis of ME etiology.