JMIR Publications

ABSTRACT

Background:

Young children vision screening, part of the preventative healthcare service, produces great value for developing regions. Besides yielding high return on investment from forestalling surgeries with low cost intervention at young ages, it improves school performance thus boosts future labor force quality. Leveraging low-skilled healthcare workers with smartphones and automated diagnosis to offer such programs can be a scalable model in resource-limited areas.

Objective:

This study aims to develop and evaluate an effective, efficient and comprehensive vision screening solution for school children in resource-limited areas. Firstly, the exams must cover major risk factors of amblyopia(弱视) and myopia, two major source of vision impairment effectively preventable at young age. Secondly, it must be integrated with digital patient record keeping for long term monitoring and popular statistical analysis. Lastly, it should utilize low-skilled technicians and only low-cost tools that are available in a typical school in developing regions, without compromising quality and efficiency.

Methods:

A workflow for the screening program is designed and a smartphone app is developed to implement it. In the standardized screening procedure, a young child goes through the smartphone-based photo-screening in a dark room. He holds a smartphone in front of his forehead, displaying pre-entered personal information as a quick response code that duplexes as a reference of scale. In one 10-second procedure, the child’s personal information and his interpupillary distance, relative visual axis alignment and refractive error ranges would be measured and analyzed automatically with image processing and artificial intelligence algorithms. His risk of strabismus, myopia and anisometropia can then be derived and consultation given.

Results:

A preliminary evaluation of the solution is conducted alongside yearly physical exams in Luoyang, Henan, P.R. China. It covered 20 suspected strabismus students and 80 randomly selected, aged evenly between 8 and 10. Each examinee takes about 1 minute and a streamlined workflow allow three of them running in parallel. The one-shot and two-shots measurement success rate is 87% and 100%, respectively. The sensitivity and specificity of strabismus detection are respectively 0.80 and 0.98. The sensitivity and specificity of myopia detection are respectively 0.83 and 1.00. The sensitivity and specificity of anisometropia detection are respectively 0.80 and 1.00.

Conclusions:

The proposed vision screening program is effective, efficient and scalable. Compared with previously published studies on utilizing smartphone for automated Hirschberg test and/or photorefraction screening, this comprehensive solution is optimized for practicality and robustness thus is better ready-to-deploy. The evaluation validates its achievement of the design specifications.