Abstract
Purpose :
Diabetic macular edema (DME) causes retinal swelling and intra-retinal fluid cysts. Optical coherence tomography (OCT) is used clinically to detect DME and treatment response. Standard clinical algorithms cannot measure cysts but only the resultant retinal thickness. Retinal thickness itself correlates poorly with visual acuity (VA) and vision recovery after DME resolution. It is impractical to manually measure cyst volume in routine practice. For this study, we used a novel automated algorithm to automatically measure DME cyst volume. We investigated correlations with the cyst volume and presenting VA or VA change in patients presenting with DME that resolved with treatment.
Methods :
The OCT scans of 40 subjects were analyzed at baseline and months 1, 3, 6, and 8. Subjects presented with DME that resolved over 8 months. A central volume was defined as a 1500 um box around the fovea, and a novel algorithm detected cysts in this central volume. The algorithm defines retinal boundaries using a graph shortest path method and uses a neutrosophic domain clustering method to detect cysts and calculate central retinal volume (CRV) and central cyst volume (CCV). Central tissue volume (CTV) was defined as CRV - CCV. Figure (1) shows OCT scans with detected retinal boundaries and cyst volumes. Pearson’s test was used to calculate pairwise linear correlation coefficients between initial CRV, CCV, and CTV with both the initial LogMAR VA and the final change in LogMAR VA at 8 months.
Results :
Results of Pearson's correlation test for initial LogMAR VA with the retinal measurements are: CRV (ρ=22.6%), CTV (ρ=30.3%), CCV (ρ=2.6%). For change in LogMAR VA the correlations are: CRV (ρ=18.1%), CTV (ρ=53.5%), and CCV (ρ=-17.1%).
Conclusions :
Initial central tissue volume correlated much better with initial VA and VA change than did central cyst volume and central retinal volume. Tissue volume allows one to distinguish the case of retinal cysts with otherwise healthy remaining tissue from the case where the surrounding tissue is severely atrophic. This may prove to be a useful biomarker for predicting patient visual prognosis prior to initiating treatment for DME.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.