Odysight: A Mobile Medical Application Designed for Remote Monitoring-A Prospective Study Comparison with Standard Clinical Eye Tests

Abstract

Introduction: The purpose of this study (Tilak Study No: TIL-001) was to evaluate the medical modules on the mobile medical application OdySight and compare them to the gold standard tests for visual acuity, contrast sensitivity, and Amsler Grid.

Methods: A total of 120 eyes were evaluated in an open-label, single-arm, prospective, single-site study during which participants performed monocular, gold standard tests for measuring visual acuity (Sloan Early Treatment Diabetic Retinopathy Study [ETDRS] letter chart at 40 cm testing distance and ETDRS letter chart at 4 m testing distance [40-cm and 4-m ETDRS, respectively), contrast sensitivity (Pelli-Robson contrast sensitivity chart [Pelli-Robson test]), and metamorphopsia/scotoma (Amsler Grid) followed by the respective modules on OdySight (also monocular). During this study, both the distance between the device and the patient's eye and room illumination were controlled by the examiner.

Results: A Bland-Altman analysis demonstrated that there was no disagreement between the results of the OdySight visual acuity module and both the 40-cm Sloan ETDRS and 4-m ETDRS tests, with a very low level of bias (0.53 and - 1.53 letters, respectively). The same analysis of contrast sensitivity showed a broader disagreement between the results of the OdySight module and those of the Pelli-Robson test. A McNemar test indicated that there was no significant difference between results obtained by the OdySight Amsler Grid module and those obtained by the paper version for the detection of metamorphopsia and scotoma (p = 1.0 for both).

Conclusion: The results from the TIL-001 study demonstrate good agreement, overall, between the measurements taken by the near visual acuity module and the Amsler grid module of OdySight as compared to currently used gold standards. The contrast sensitivity module of OdySight will require additional investigation. OdySight could be used for remote monitoring of vision between clinic visits and potentially assist in follow-up planning.

Funding: Tilak Healthcare funded the study and the Rapid Service Fees.

Trial registration: ClinicalTrials.gov identifier: NCT03457441.

Keywords: AMD; Amsler grid; App; ETDRS; Eye diseases; Mobile medical game; OdySight; Remote monitoring; Visual acuity; mHealth.

Figures

Fig. 1

OdySight. a Visual test, b game, c physician’s dashboard. Patients play the game and perform the tests. Physicians can follow their patients using the online dashboard

Fig. 2

OdySight visual acuity module (the hand and arrow are for comprehension only)

Fig. 3

OdySight contrast sensitivity module (the hand and arrow are for comprehension only)

Fig. 4

Three screens depicting the digital Amsler grid on OdySight

Fig. 5

Participant disposition. ITT Intention to treat

Fig. 6

Bland and Altman (B&A) plots comparing absolute differences between visual acuity measured (OdySight vs. Sloan Early Treatment Diabetic Retinopathy Study [ETDRS] letter chart at 40 cm testing distance [40cm Sloan ETDRS]) and the mean visual acuity measured ([40cm Sloan ETDRS + OdySight)]2). For an explanation of cohorts, please see footnote to Table 1. LoA Limits of agreement, s standard deviation

Fig. 7

B&A plots comparing absolute differences between visual acuity measured (OdySight vs, 4-m ETDRS) and the mean visual acuity measured ([4-m ETDRS + OdySight]/2)

Fig. 8

B&A plots comparing absolute differences between contrast sensitivity measured (OdySight vs. Pelli–Robson Contrast Sensitivity Chart) and the mean contrast sensitivity measured ([Pelli-Robson + OdySight]/2)

Fig. 9

Example of corresponding results for two versions of the Amsler grid. Left: paper 10 × 10-cm Amsler grid completed with a felt tip pen. Right: digital Amsler grid completed on the OdySight app