Innovative Imaging Discovers Early Retinal Changes in Type 2 Diabetes

In a new study, texture analysis of optical coherence tomography (OCT) images identified structural retinal changes before clinically apparent diabetic retinopathy in early type 2 diabetes.

The investigators ran a repeated-measures experimental cohort with OCT and electroretinography at baseline and at 4, 8, and 12 weeks, applying automated segmentation to derive layer-specific thickness and texture metrics. Primary endpoints were layer-level texture features and OCT thickness; texture metrics began to diverge from controls at weeks 8-12, underscoring their diagnostic potential for earlier detection.

Layer-level findings included thinning of the inner plexiform layer and the inner/outer photoreceptor segments, together with altered macular texture measures—autocorrelation, cluster prominence, correlation, homogeneity, information measure of correlation II, inverse difference moment normalized, and sum average. These signals form a set of candidate retinal biomarkers for subclinical disease; focused layer analysis therefore provides specific metrics for validation studies.