Software-assisted analysis during ocular health screening

Abstract

Objective: To describe the use of software-assisted analysis of non-mydriatic 45 color fundus images during comprehensive ocular screening. Materials and Methods: Software-driven filters (blue [490 nm; nerve fiber layer], green [550 nm; neural retina], and red [610 nm; pigmented retina/choroid]) and an "emboss" (topographic changes) digital filter were used to enhance image analysis during ocular health screening performed using a Canon (Tokyo, Japan) 8.2-megapixel non-mydriatic retinal camera. Intraocular pressure (model TX-f full auto non-contact tonometer; Canon) and visual acuity (SIMAV, Padova, Italy) were also determined. An on-site medical director analyzed the collected data, provided immediate subject feedback, and made recommendations and referrals for general or specialty ophthalmology clinics. Results: Software-assisted analysis was performed on color images from 128 veterans (mean age, 37.5 years; 66% male; 43% white). Software filters allowed efficient image analysis at a rate of 26 eyes/h. Thirteen (10.2%) persons had a finding consistent with a vision-threatening disease and were referred for a dilated fundus examination. Conclusions: Software-assisted screening permits thorough and efficient evaluation of ocular health during an ocular screening event. © Mary Ann Liebert, Inc.