A Deep Learning Approach for Cataract Detection Using VGG16

Abstract

Cataracts are a leading cause of visual impairment globally, especially among the aging population. Early and accurate detection is essential to prevent irreversible vision loss. This study proposes a deep learning–based approach for automated cataract detection using retinal fundus images. The VGG16 convolutional neural network (CNN) was employed and fine-tuned for binary classification between normal and cataract-affected eyes. The dataset, sourced from publicly available repositories, was limited to two classes, Normal and Cataract—with approximately 1,000 images per class. Data preprocessing and augmentation techniques were applied to enhance model generalization and robustness. The model achieved an accuracy of 92%, indicating its reliability for clinical use. Comparative analysis with existing literature demonstrates the model’s competitive performance, highlighting its potential as a practical and accessible tool for aiding ophthalmologists in early cataract diagnosis. Future work may include multi-disease detection and the integration of explainable AI for increased transparency in medical decision-making.

Article Content

Keywords

Article Info