Efficient Mask Detection and Stage Classification of Diabetic Retinopathy Using Lightweight CNN and U-Net Models
International Journal of Drug Delivery Technology · 29.06.2026 · 16 (59s)
DOI: 10.25258/ijddt.16.59s.100 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Drug Delivery Technology, online erschienen am 29.06.2026. Verfasst von Kajal Kadusing Patil und 4 weiteren Autor:innen.
Abstract (Originalsprache)
Effective treatment of diabetic retinopathy (DR), a major cause of vision impairment worldwide, depends on early identification. By using lightweight Convolutional Neural Networks (CNNs) and a customized U-Net architecture, this study offers an effective deep learning framework for automated classification and mask detection of diabetic retinopathy. The main goal is to maximize computing economy without sacrificing classification accuracy so that the solution can be used in clinical settings with limited resources. The suggested approach initially uses a classification technique such as CNN, MobileNetV2 and Proposed MobileNetV2 architecture to classify diabetic retinopathy images and then Efficient U-Net model is applied to accurately segment retinal lesions and anatomical features like exudates and microaneurysms. Publicly accessible (IDRiD) datasets were used in extensive trials, and performance was assessed using measures including as accuracy, F1-score, sensitivity, and processing cost. The findings show that, in comparison to conventional deep models, the suggested framework greatly reduces model complexity and inference time while achieving strong classification performance of 99% accuracy and segmentation accuracy of 98%. All things considered, a viable strategy for effective, precise, and scalable DR screening and monitoring systems is the combination of lightweight MobileNetV2 and Efficient U-Net models.
Zitieren
APA
Patil, K. K., Chiddarwar, G., Thakre, K. S., Chaudhari, S., & Shekapure, S. (2026). Efficient Mask Detection and Stage Classification of Diabetic Retinopathy Using Lightweight CNN and U-Net Models. International Journal of Drug Delivery Technology, 16(59s). https://doi.org/10.25258/ijddt.16.59s.100
Vancouver
Patil KK, Chiddarwar G, Thakre KS, Chaudhari S, Shekapure S. Efficient Mask Detection and Stage Classification of Diabetic Retinopathy Using Lightweight CNN and U-Net Models. International Journal of Drug Delivery Technology. 2026;16(59s). doi:10.25258/ijddt.16.59s.100
BibTeX
@article{Patil2026,
author = {Patil, Kajal Kadusing and Chiddarwar, Girija and Thakre, Kalpana sunil and Chaudhari, Smita and Shekapure, Swati},
title = {Efficient Mask Detection and Stage Classification of Diabetic Retinopathy Using Lightweight CNN and U-Net Models},
journal = {International Journal of Drug Delivery Technology},
year = {2026},
volume = {16},
number = {59s},
doi = {10.25258/ijddt.16.59s.100}
}
EUVIMED