Artificial intelligence-based quantification of retinal microvascular biomarkers from fundus photography of chronic kidney disease: a case-control study

1

Retinal microvascular alterations may serve as non-invasive indicators for early detection of chronic kidney disease (CKD).
2

AI-based analysis of fundus images identified specific retinal parameters associated with CKD, including arteriovenous ratio and vascular density.
3

The study analyzed 322 participants, revealing that certain retinal features were altered even in early-stage CKD.
4

The final model demonstrated moderate discrimination for CKD with an area under the curve (AUC) of 0.776.
5

Further validation of AI-derived retinal microvascular parameters in diverse populations is necessary to confirm their clinical utility.

Frontiers In Medicine

by Qiumei Gu, Min Liu, Weiwei Zhang, Zhengju Chen, Xingye Wang, Jie Wang, Ziyan He, Fang Lu
May 18, 2026

1