Imagine losing your sight before you even know something’s wrong. That’s the grim reality for millions living with diabetes, as diabetic retinopathy (DR) silently damages their eyes, often undetected until it’s too late. But what if we could catch these changes before they steal vision? A groundbreaking study from the University of Coimbra, Portugal, has unlocked a new way to spot early retinal trouble using a technique called texture-based analysis of OCT images. And this is the part most people miss: it’s not just about seeing clearer pictures—it’s about decoding the hidden patterns within those images that signal trouble long before traditional methods can.
Diabetic retinopathy is a leading cause of blindness among working-age adults, affecting over 130 million people globally. While advancements in eye imaging have improved monitoring, most patients are diagnosed only after years of unnoticed retinal damage. Early changes at the molecular and cellular levels—like neurodegeneration, inflammation, and vascular dysfunction—often slip past standard tests like fundus photography or angiography. As a result, irreversible vision loss can occur before anyone even knows there’s a problem. This gap in early detection has left doctors and patients scrambling for better solutions.
Enter the University of Coimbra’s research team, who’ve developed a texture-based analysis of optical coherence tomography (OCT) images that can detect subtle retinal changes in type 2 diabetes. Published in Eye and Vision on September 3, 2025, the study used a rat model mimicking diabetes to track retinal alterations over 12 weeks. By measuring microscopic texture variations, the method uncovered neurovascular abnormalities far earlier than traditional biomarkers or signs of vascular leakage.
Here’s how it works: the researchers analyzed over 80 retinal scans from diabetic and healthy rats, using a gray-level co-occurrence matrix (GLCM) to quantify texture parameters across retinal layers. Out of 20 features examined, eight—including autocorrelation, cluster prominence, correlation, homogeneity, information measure of correlation II (IMCII), inverse difference moment normalized (IDN), inverse difference normalized (INN), and sum average—showed significant changes in diabetic retinas, especially in the inner plexiform layer (IPL) and photoreceptor segments (IS/OS). But here’s where it gets controversial: seven of these metrics had also been flagged in a previous study using a type 1 diabetes model, raising the question—could these texture changes be a universal early warning sign for all types of diabetes? Despite minimal thinning and delayed oscillatory potentials, the retinas showed no major inflammation or vascular leakage, confirming that texture changes are the canary in the coal mine for retinal damage.
These findings position texture analysis as a powerful tool for detecting early structural disorganization in the retina, potentially closing the gap between biological changes and clinical diagnosis. Professor António Francisco Ambrósio, co-senior author of the study, emphasizes, ‘Our results demonstrate that texture analysis can uncover minute retinal changes long before DR becomes clinically visible. By capturing subtle structural signals within OCT images, this approach opens a new diagnostic window into the earliest disease processes.’
The implications are huge. This method could help identify high-risk patients before permanent vision damage occurs, enabling earlier treatment and better outcomes. The consistency of these texture metrics across diabetes models suggests they could serve as universal early biomarkers. But is this the silver bullet we’ve been waiting for? While promising, further clinical trials are needed to validate these findings in humans and refine algorithms for large-scale screening and teleophthalmology applications.
This research lays the foundation for AI-assisted diagnostic tools that could automatically screen for preclinical DR based on retinal texture signatures. Integrating this analysis into routine OCT imaging could allow ophthalmologists to spot microscopic structural disruptions—even in patients with seemingly normal vision. Such early detection could revolutionize personalized care, prevent irreversible retinal damage, and slash the global burden of diabetic blindness.
So, here’s the big question: Could texture-based OCT analysis become the new gold standard for early diabetes-related eye care? And if so, how quickly can we bring this technology to clinics worldwide? Share your thoughts in the comments—let’s spark a conversation that could shape the future of eye health.
