Innovative myopic screening platform based on smartphones

Qiang Su¹Yicheng Ge²Jinghui Wang¹Nan Jin¹Haochen Han¹Yingxin Li³Chea Su Kee⁴Bei Du^1*Ruihua Wei^1*

• ¹Tianjin Medical University Eye Hospital, Tianjin, China
• ²Department of Ophthalmology, Shanghai Ninth People’s Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China
• ³Cancer Hospital Chinese Academy of Medical Sciences, Beijing, China
• ⁴The Hong Kong Polytechnic University, Hong Kong, Hong Kong, SAR China

Introduction: To validate a novel smartphone-based approach for subjective refraction, specifically for myopia screening, offering a cost-effective and accessible tool for at-home vision assessment. Methods: A total of 230 healthy volunteers, encompassing 460 eyes, aged between 7 and 40 years (mean±SD: 21.04±7.76), and exhibiting refractive error (RE) ranging from -6.25 to +0.50 diopter (D), were deemed eligible subjects in this research. Subjective refraction assessments were conducted on all subjects, involving both the conventional phoropter examination by experienced optometrists which served as the clinical gold-standard, and the smartphone-based methodology. During smartphone measurement, the screen was gradually moved toward the eye until achieving clear differentiation of the 'E' optotypes and the astigmatic dial. The eye-to-smartphone screen distances (ESD) were calculated based on the image of iris diameter (ID) acquired by the front-facing camera of the smartphone. Applying the definitions of the far point and the rule of thirty, this allowed for the computation of sphere (S), cylinder (C), and astigmatism axis (α) values. The concordance between the two methods was assessed by establishing the Limits of Agreement (LOAs), which were calculated as the mean difference ± 1.96 times the standard deviation of the differences. Results: The smartphone‑based screening technique showed that it closely matched the gold-standard subjective refraction used in clinical practice. The LOAs for S, C, and spherical equivalent refraction (SER) were 0.11±0.89 D, -0.03±0.82 D, and 0.10±0.89 D, respectively. The average deviation in measuring the α was 4.72°, with 64.35% of deviations falling within the ±15° range. Additionally, the technique demonstrated impressive areas under the receiver operating characteristic curve (AUC) of 0.973 for the range of -3.00 D < SER ≤ 0.00 D and 0.986 for the range of - 6.00 D < SER ≤ -3.00 D. Discussion: The innovative vision screening through smartphones can expand access to measuring RE, especially in home. The study confirms the validity of an innovative vision-screening approach for myopia. Its effectiveness and accessibility make it a valuable tool for opportunistic or large-scale myopia screening programs.