Abstract
Purpose :
To evaluate patients with diabetic retinopathy using ultra-widefield fluorescein angiography (UWFA), and to use a novel medium to dynamically visualize serial UWFA imaging to observe the clinical value of UWFA dynamic video in diabetic retinopathy diagnosis.
Methods :
In this observational study, 23 eyes of 23 patients with diabetic retinopathy were evaluated at a single institution underwent UWFA. Capillary non-perfusion area, intraretinal microvascular abnormality (IRMA) and retinal neovascularisation (NV) was compared to UWFA to the inlaid Early Treatment of Diabetic Retinopathy Study (ETDRS) 7-field template. Each serial images were composed into a time-lapse video in a Moving Picture Experts Group 4 (MP4) movie format and characteristic of NVs in UWFA were analyzed.
Results :
The DR severity included: mild nonproliferative DR (NPDR), 17.4% (4 eyes); moderate NPDR, 17.4% (4 eyes); severe NPDR, 17.4% (4 eyes); proliferative DR (PDR), 47.8% (11 eyes). UWFA was able to capture 3.9 times more area of retinal surface than ETDRS (P<0.001), which was also capable of demonstrated 2.6 times (P=0.005) as much area of IRMA, 8.7 times (P=0,003) times as much area of non-perfusion area, 1.9 times (P=0.02) as much area of NV as in ETDRS. The average number of images captured in each eye was 66 and the video run 9.7 seconds in average. There was no significant difference between the average non-perfusion index of UWFA traditional imaging group and dynamic imaging group. The retinal neovascularisation leakage area increased linearly with time (r=0.996) which was 75% times greater by 2 minutes and 189% greater by 5 minutes.
Conclusions :
Compared with ETDRS imaging, UWFA imaging showed more details of peripheral retinal pathology which was of great importance to evaluate the DR severity. The UWFA dynamic video could show the whole process of angiography, the novel medium was helpful to identify the confusing lesions and beneficial to academic communication, clinical education and the doctor-patient communication.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.