Topology-preserving, diameter-specific framework for conjunctival vessel segmentation and tortuosity analysis in diabetes.
Analyzing eye vessel patterns from simple external photos can detect microscopic blood vessel damage in diabetes without expensive equipment, potentially enabling screening in resource-limited settings.
A novel topology-preserving deep learning framework enables accurate conjunctival vessel segmentation and tortuosity quantification in diabetes, demonstrating significantly elevated vessel tortuosity versus healthy controls using only external eye imaging. Conjunctival imaging may offer a scalable, non-invasive, slit-lamp-free tool for systemic microvascular phenotyping in diabetes.
What the study was
- Study design
- Prospective observational + ML model development
- Population
- Diabetes patients and healthy controls (Sri Lanka)
- Category
- Diagnostics
- Maturity
- Exploratory
- Journal
- Scientific Reports
Why it surfaced
Novel non-invasive microvascular imaging tool for diabetes; Sci Rep publication; validated against retinal findings; N and demographic details missing from abstract; Sri Lanka single-center limits generalizability.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.