Remote ophthalmology with a smartphone adapter handled by nurses for the diagnosis of eye posterior pole pathologies during the COVID-19 pandemic

Abstract

Introduction: The use of smartphones to provide specialist ophthalmology services is becoming a more commonly used method to support patients with eye pathologies. During the COVID-19 pandemic, demand for telehealth services such as tele-ophthalmology, is increasing rapidly. Methods: In 2019, the agreement between diagnostic tests was investigated by comparing the diagnostic performance for eye posterior pole pathologies of the images obtained by a smartphone coupled to a medical device known as open retinoscope (OR), handled by a nurse and subsequently assessed by an ophthalmologist versus the images obtained by an ophthalmologist using a slit lamp associated to a 76 diopter indirect ophthalmic lens (Volk Super FieldVR ) (SL-IOL) at the outpatient department of a hospital. The OR used in this study worked with a 28 diopter indirect lens. Results: An examination of 151 dilated eyes (79 adult patients, mean age of 66.7 years, 59.5% women) was conducted. Sensitivity was 98.9%, specificity was 89.8%, the positive predictive value was 93.8% and the negative predictive value was 98.2%. The kappa index between both tests was 0.90 (95% CI: 0.83-0.97) in basic diagnosis, 0.81 (95% CI: 0.74-0.89) in syndromic diagnosis (13 categories) and 0.70 (95% CI: 0.62-0.77) in advanced diagnosis (23 categories). Discussion: Images obtained by a nurse using a smartphone coupled to the OR and subsequently assessed by an ophthalmologist showed a high diagnostic performance for eye posterior pole pathologies, which could pave the way for remote ophthalmology systems for this patient group.

Literature Review on Artificial Intelligence Methods for Glaucoma Screening, Segmentation, and Classification

“…The multiplicity of conditions under which photos are taken in different places using cameras with different settings, taken from various groups of patients, with objectives defined by multiple specialists, can interfere with the results [52]. The databases for the study of pathologies of the posterior pole of the eye show the retina, vessels, macula, and optic disc [53]. Other image databases for the study of glaucoma centered on the optic papilla with demarcations of the papilla and excavation boundaries made by specialists represent the gold standard, serve as reference standards for segmentation measurements, and can exploit the implementation of DL networks [54].…”