Scholarly open access journals, Peer-reviewed, and Refereed Journals, Impact factor 8.14 (Calculate by google scholar and Semantic Scholar | AI-Powered Research Tool) , Multidisciplinary, Monthly, Indexing in all major database & Metadata, Citation Generator, Digital Object Identifier(DOI)
Every two seconds, somewhere in the world, di-abetes silently begins destroying someone’s eyesight. Dia- betic retinopathy (DR)—a progressive disease of the retinal blood vessels—is responsible for millions of cases of pre- ventable blindness, yet it announces itself with no pain and no warning until irreversible damage is already done. The only defense is early, systematic screening. But with hun- dreds of millions of diabetic patients and a global shortage of ophthalmologists, manual screening at scale is simply impossible.
This paper tells the story of how we built a solu- tion. We designed and trained a deep learning system centered on ConvNeXt—a next-generation convolutional architecture—augmented with channel-wise attention, a flex- ible Kolmogorov-Arnold-inspired classifier, and a consis- tency regularization training regime. The result is a binary diagnostic model that, on the APTOS 2019 benchmark, cor- rectly identifies the presence or absence of diabetic retinopa- thy 98.56% of the time—surpassing every competing single- model and ensemble approach published to date on the same dataset. More than a number, this represents a system gen- uinely ready to stand alongside clinicians, catch the cases that would otherwise be missed, and protect the vision of patients who cannot afford to wait.
Keywords:
Diabetic retinopathy, retinal fundus images, deep learning, ConvNeXt, squeeze-and-excitation attention, Kolmogorov-Arnold networks, consistency regularization, medical decision-support systems
Cite Article:
"Diabetic Retinopathy Detection using ConvNeXt: Advancing Early Vision Loss Prevention with Deep Learning", International Journal for Research Trends and Innovation (www.ijrti.org), ISSN:2456-3315, Vol.11, Issue 3, page no.a151-a158, March-2026, Available :http://www.ijrti.org/papers/IJRTI2603025.pdf
Downloads:
000144
ISSN:
2456-3315 | IMPACT FACTOR: 8.14 Calculated By Google Scholar| ESTD YEAR: 2016
An International Scholarly Open Access Journal, Peer-Reviewed, Refereed Journal Impact Factor 8.14 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage Journal Indexing in All Major Database & Metadata, Citation Generator