Abstract: Diabetic Retinopathy (DR) is a serious consequence of diabetes mellitus that can result in irreversible vision loss if not discovered and treated promptly. Traditional manual diagnosis by ophthalmologists is labor-intensive, time-consuming, and error-prone. In recent years, deep learning approaches, particularly Convolutional Neural Networks (CNNs) like Xception, EfficientNet, and DenseNet, have demonstrated extraordinary efficacy in medical image analysis, including DR detection and categorization. This project covers and analyzes cutting-edge approaches for detecting and categorizing DR in color fundus pictures using deep learning techniques. These technologies show great promise for automating and streamlining the DR screening process, thereby lowering costs and increasing efficiency in healthcare delivery. The necessity of regular retina screening for diabetic people cannot be emphasized, as the risk of DR grows as diabetes progresses. Early diagnosis of DR lesions, such as microaneurysms, hemorrhages, and exudates, is critical for timely treatments and preventing vision loss. Computer-aided diagnosis systems that use deep learning algorithms have the potential to transform DR screening by giving accurate and fast assessments, allowing for rapid treatment and lowering the risk of blindness among diabetics.
Keywords: Xception, Densenet, Efficientnet, microaneurysms, hemorrhages
Cite:
Dr.V. Ramachandran, Akhila Patchala, Lakshmi Sowjanya Potla,Phinehas Prakash Jupudi, Rohith Sai Obilisetty, "Diabetic Retinopathy Detection using Deep Learning Techniques", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 2, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13252.
|
DOI:
10.17148/IJARCCE.2024.13252
