Artificial intelligence-based analysis of retinal vascular changes in the preclinical and early stages of diabetic retinopathy using ultra-widefield fundus imaging: an observational cross-sectional study.
Artificial intelligence analysis of retinal vessels distinguishes early diabetic changes from healthy eyes.
Artificial intelligence-based analysis of retinal vascular changes in the preclinical and early stages of diabetic retinopathy using ultra-widefield fundus imaging: an observational cross-sectional study.
Subclinical retinal microvascular remodeling may occur before clinically detectable diabetic retinopathy (diabetic retinopathy).
This study investigated artificial intelligence (artificial intelligence)-derived ultra-widefield retinal vascular metrics in patients with type 2 diabetes mellitus (T2DM) with and without non-proliferative diabetic retinopathy (non-proliferative DR), and evaluated their potential for early vascular phenotyping and diagnostic discrimination.
In this observational cross-sectional single-center study, 237 participants were included: 63 healthy controls (103 eyes), 101 patients with T2DM without diabetic retinopathy (No-diabetic retinopathy; 201 eyes), and 73 patients with non-proliferative DR (132 eyes).
Artificial intelligence test correctly told diabetic from healthy eyes most of the time
After quality control, covariate adjustment, and false discovery rate correction, 38 vascular-parameter rows remained significant.
Whole-field vessel density was highest in the No-DR group, intermediate in non-proliferative DR, and lowest in controls [control, 0.017 (0.008-0.023); No-DR, 0.026 (0.020-0.032); non-proliferative DR, 0.021 (0.015-0.026); false discovery rate P < 0.001].
Artificial intelligence-derived ultra-widefield retinal vascular metrics demonstrate early, spatially heterogeneous microvascular remodeling in T2DM before clinically apparent diabetic retinopathy.
Vessel density, fractal dimension, vessel diameter, and vessel length provide complementary information, and multiparameter vascular modeling may support early detection and risk stratification.
External validation and longitudinal studies are required before clinical implementation.