Joslin Vision Network Validation Study: pilot image stabilization phase

Abstract

Purpose: This preliminary study is designed to evaluate the quality of initial images obtained using the Joslin Vision Network (JVN) image capture and retrieval system.

Methods: Digitized images of 18 patients (36 eyes) were obtained using Topcon Non-mydriatic Fundus Camera and various imaging protocols as described for each series of images. Level of retinopathy ranged from no diabetic retinopathy to proliferative diabetic retinopathy. Images were reviewed and evaluated in comparison to medical record notes and Early Treatment Diabetic Retinopathy Study (ETDRS) seven-standard field stereoscopic 35-mm retinal fundus photographs--when available--to determine the ability of JVN digitized images to allow appropriate diagnosis and clinical management. JVN images were also evaluated for presence of photographic artifacts and stereoscopic presentation. The various displays and relative value of each JVN display were also evaluated.

Results: Using combinations of JVN image displays, JVN images were shown to match diagnosed level of diabetic retinopathy as determined from record notes and/or ETDRS seven-standard field stereoscopic 35-mm retinal fundus photography. Viewing images with various combinations of JVN image displays delineates possible sources of artifacts in JVN images. Areas to guide future image capture and image reading are identified.

Conclusions: The JVN image capture system and reading of images by certified readers should allow appropriate photodiagnostic information to support the JVN, the diabetes mellitus disease management telemedicine initiative of the Joslin Diabetes Center. A rigid, full-scale, masked study to compare the Joslin Vision Network to seven-standard field stereoscopic 35-mm retinal fundus photography should be pursued to validate the image-gathering and image-reading strategy to support the Joslin Vision Network.