Abstract
Purpose :
Hallmarks of sickle cell retinopathy (SCR) include sea fanning neovascularization, salmon patches, and non-perfused areas. To date, no studies have looked to quantitatively define SCR lesions and their change in response to treatment. Herein we seek to characterize a novel approach in quantifying SCR on fundus angiograms.
Methods :
Wide spectrum fluorescence angiography (WSFA) scans were accessed at the Kresge Eye Institute, Detroit, MI from 2005 to present. Those patients that were 18 years or older and had a pre and post treatment WFSA were included. Patients were excluded from the study if they had received treatment for SCR outside of the study window. Age, sex, eye treated, treatment date, and visual acuity (VA) before and after treatment were extracted from the patient charts. Lesions of sea fanning neovascularization were manually mapped out by raters. Lesion area and intensity on WSFA scans were analyzed using ImageJ software Fiji plugin. ImageJ is an open source image-processing package that has been previously applied to ophthalmic imaging. Two readers individually assessed the images two times. Pixel number and lesion intensity were documented. Area and intensity were standardized against the optic disc area and average intensity for each image.
Results :
Eleven patients met the inclusion and exclusion criteria. Of these patients, 5 (45.5%) were female and 6 (54.5%) male. Mean age at the time of treatment was 46.5 ± 16.3 years. Prior to treatment, the mean lesion area as a multiple of disc area was 6.24 ± 5.54 while intensity was 33.2 ± 54.4. Post-treatment mean lesion area was 2.57 ± 2.62 disc area while intensity was 10.2 ± 88.0. The change in lesion area was significant (p=0.009) but the intensity change was not significant (p=0.297). Intra and inter rater reliabilities of intensity and area measures were highly correlated with all measures exceeding 0.88 and significant (p<0.05).
Conclusions :
At present, clinicians are limited in their ability to quantify SCR which makes defining treatment success difficult. In this investigation we sought to quantify SCR pre-to-post treatment by assessing WSFA documented lesion area and intensity. Using imageJ analysis we demonstrated the change in lesion area was significant while maintaining a high level of intra and inter rater reliability. These results suggest imageJ may be highly sensitive in capturing disease regression.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.