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

IF 2.7 2区医学 Q1 OPHTHALMOLOGY

Experimental eye research Pub Date : 2025-08-30 DOI:10.1016/j.exer.2025.110619

Amal Alzu'bi , Sondos Momany , Abdelwahab Aleshawi , Mais Tashtoush , Rami Al-Dwairi

{"title":"The utility of artificial intelligence in characterization and detecting causes of macular edema: A spectral-domain OCT-based algorithm study","authors":"Amal Alzu'bi , Sondos Momany , Abdelwahab Aleshawi , Mais Tashtoush , Rami Al-Dwairi","doi":"10.1016/j.exer.2025.110619","DOIUrl":null,"url":null,"abstract":"<div><h3>Background</h3><div>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.</div></div><div><h3>Methods</h3><div>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.</div></div><div><h3>Results</h3><div>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.</div></div><div><h3>Conclusions</h3><div>The findings support the integration of a robust CNN architecture and XAI techniques to enhance diagnostic precision and aid clinical decision-making in ophthalmology.</div></div>","PeriodicalId":12177,"journal":{"name":"Experimental eye research","volume":"260 ","pages":"Article 110619"},"PeriodicalIF":2.7000,"publicationDate":"2025-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Experimental eye research","FirstCategoryId":"3","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0014483525003902","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"OPHTHALMOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Background

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.

Methods

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.

Results

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.

Conclusions

The findings support the integration of a robust CNN architecture and XAI techniques to enhance diagnostic precision and aid clinical decision-making in ophthalmology.

查看原文本刊更多论文

人工智能在黄斑水肿表征和病因检测中的应用：基于光谱域oct的算法研究

黄斑水肿（ME）是一种常见的视力丧失原因，可由各种视网膜疾病引起，最显著的是糖尿病性黄斑水肿（DME）和年龄相关性黄斑变性（AMD）。准确和及时地区分这些原因对于适当的治疗是必要的；然而，它仍然是一个诊断挑战。本研究通过开发和评估一个深度学习框架来解决自动化ME分类的空白，该框架能够使用光学相干断层扫描（OCT）图像区分DME、AMD和正常视网膜状况。方法采用回顾性数据集，包括来自阿卜杜拉国王大学医院（KAUH）的1040张OCT图像，并结合公共数据集进行基准测试。将数据集分为带注释和未带注释的图像，通过预处理、增强和模拟分割来提高模型性能。我们对三个预训练的卷积神经网络resnet152、InceptionV3和MobileNetV2进行了基准测试和评估。结果InceptionV3和ResNet152在两个数据集上的准确率最高（95% - 98%）。另一方面，MobileNetV2在KAUH数据集上表现出中等的准确性（89%），但在公共数据集上表现出很强的性能（97%）。可解释人工智能（XAI）技术，特别是Grad-CAM，被应用于可视化模型预测，并根据注释数据手动验证结果，以评估可解释性。结论：该研究结果支持强大的CNN架构和XAI技术的整合，以提高眼科的诊断精度和辅助临床决策。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Experimental eye research 医学-眼科学

CiteScore

6.80

自引率

5.90%

发文量

323

审稿时长

66 days

期刊介绍： The primary goal of Experimental Eye Research is to publish original research papers on all aspects of experimental biology of the eye and ocular tissues that seek to define the mechanisms of normal function and/or disease. Studies of ocular tissues that encompass the disciplines of cell biology, developmental biology, genetics, molecular biology, physiology, biochemistry, biophysics, immunology or microbiology are most welcomed. Manuscripts that are purely clinical or in a surgical area of ophthalmology are not appropriate for submission to Experimental Eye Research and if received will be returned without review.