ASSESSMENT OF A SMARTPHONE-BASED IMAGING SYSTEM FOR THE DETECTION OF CORNEAL DISEASES

The development of optical adaptors such as the QuikVue Plus system allows clinicians to capture high-quality anterior segment images using most available smartphones. This study aimed to evaluate the diagnostic accuracy, image quality, and interobserver reliability of this smartphone-based anterior segment imaging adapter compared to conventional slit-lamp photography under routine clinical conditions.

Patients were recruited in the corneal and emergency outpatient clinic of the Department of Ophthalmology and Optometry, Kepler University Clinic 
Kepler University, Linz, Austria

Between April 2024 and November 2024, a total of 101 patients presenting to the corneal outpatient clinic and emergency department at Kepler University Hospital Linz were screened and 99 patients were enrolled for analysis. Each patient underwent anterior segment imaging using both the QuikVue Plus (Visuscience Meditech Co., LTD, China). smartphone adapter and a conventional slit lamp during the same visit. Images were independently assessed by one corneal specialist and two ophthalmology residents. Diagnostic accuracy, agreement between modalities and observers were analyzed using Cochran´s Q test, McNemar´s test and t-test. Image quality was evaluated between observers using a five-point Likert scale.

Mean age of the study cohort was 59.28 years. Of the 99 patients, 57 were female and 42 were male, resulting in a nearly balanced sex distribution.

The corneal specialist achieved diagnostic agreement between devices in 78.8% of all images, indicating no statistically significant difference between modalities (p = 0.45). Individually he could diagnose 88.9% of pathologies correctly using slit lamp images, comparable to 84.9% of correct diagnoses with the QuickVue Plus adapter. Residents achieved no significant agreement for the right diagnoses only between the devices. The interobserver agreement was high between residents for each device indivudually but significantly different from the specialist who outperformed both residents (p = <0.001). Both devices together showed significant difference between residents (p = 0.048 adj.) and no significant difference between the specialist and resident 1 after correction. Image quality ratings showed a high percentage difference between devices based on the evaluater.

The QuikVue Plus smartphone adaptor enabled anterior segment imaging with diagnostic accuracy slightly lower but comparable to conventional slit-lamp photography based on the level of experience of the observer. This indicates that for a qualitative clinical image evaluation only an experienced corneal specialist can achieve high diagnostic accuracy comparable to a slit lamp. These findings support the potential of smartphone-based imaging as a reliable, low-cost, and portable diagnostic alternative for ophthalmic practice. Such systems may be especially valuable for teleophthalmology applications and eye care delivery in resource-limited or remote settings.