Abstract:

Diabetic Retinopathy (DR) is a leading cause of vision impairment among diabetic patients worldwide. Early detection is crucial for preventing severe vision loss. This study proposes DiaVision, an AI-powered system that leverages deep learning techniques for automated detection of DR from retinal fundus images. The system integrates convolutional neural networks (CNNs) for feature extraction and classification, achieving high accuracy, sensitivity, and specificity. Experimental results demonstrate that DiaVision outperforms traditional machine learning approaches, enabling rapid and reliable screening, which can significantly reduce the burden on ophthalmologists and improve patient outcomes.

Description:

.