Purpose: To illustrate a noninvasive method to analyze the retinal capillary lumen caliber in patients with Type 1 diabetes.
Methods: Adaptive optics imaging of the retinal capillaries were acquired in two parafoveal regions of interest in eyes with nonproliferative diabetic retinopathy and unaffected controls. Measures of the retinal capillary lumen caliber were quantified using an algorithm written in Matlab by an independent observer in a masked manner. Comparison of the adaptive optics images with red-free and color wide fundus retinography images was also assessed.
Results: Eight eyes with nonproliferative diabetic retinopathy (eight patients, study group), no macular edema, and preserved visual acuity and eight control eyes (eight healthy volunteers; control group) were analyzed. The repeatability of capillary lumen caliber measurements was 0.22 μm (3.5%) with the 95% confidence interval between 0.12 and 0.31 μm in the study group. It was 0.30 μm (4.1%) with the 95% confidence interval between 0.16 and 0.43 μm in the control group. The average capillary lumen caliber was significantly narrower in eyes with nonproliferative diabetic retinopathy (6.27 ± 1.63 μm) than in the control eyes (7.31 ± 1.59 μm, P = 0.002).
Conclusion: The authors demonstrated a noninvasive method to analyze, with micrometric scale of resolution, the lumen of retinal capillaries. The parafoveal capillaries were narrower in patients with Type 1 diabetes and nonproliferative diabetic retinopathy than in healthy subjects, showing the potential capability of adaptive optics imaging to detect pathologic variations of the retinal microvascular structures in vaso-occlusive diseases.
The authors described a noninvasive method to analyze the lumen caliber of retinal capillaries using adaptive optics imaging. The repeatability of the method was evaluated in a study of the parafoveal capillaries in eyes with nonproliferative diabetic retinopathy and unaffected controls.
*Fondazione G.B. Bietti IRCCS, Rome, Italy;
†CNR-IPCF Unit of Support Cosenza, University of Calabria, Rende, Italy; and
‡Vision Engineering, Rome, Italy.
Reprint requests: Marco Lombardo, MD, PhD, Fondazione G.B. Bietti IRCCS, Via Livenza 3, 00198 Rome, Italy; e-mail: firstname.lastname@example.org
None of the authors have any financial/conflicting interests to disclose.