Evaluating The Validity Of Chronosight Pilot's Automated Refraction Using An Artificial Intelligence Algorithm Compared To Standard Orthoptist Refraction Techniques

To evaluate the validity of Chronos SightPilot (Topcon Corporation, Tokyo, Japan) as compared with standard refraction by orthoptist using ARK (Nidek Technologies, Gamagori, Japan) and subjective refraction refinement.

Private clinic - Clinique Honoré Cave à Montauban, France

Participants received both objective and subjective refraction assessments using Chronos with SightPilot refraction algorithm under binocular conditions and the Nidek ARK monocularly (referred to as "standard") and subjective refraction refinement on the same day by the same operator. The results were categorized into four groups based on the difference in refraction outcomes between the Chronos and standard methods: high concordance (difference ≤0.25D), medium concordance (difference >0.25D and ≤0.50D), low concordance (difference >0.5 and ≤0.75 D) and no concordance (difference >0.75D), streamlining the classification by correcting the overlap in concordance levels.

The study included 369 eyes from 187 adult patients, with an average age of 38.1 years (±17.9 years). Concordance rates of the sphere between Chronos binocular objective refraction and standard monocular objective refraction were found to be: high 78.9%, medium 14.7%, low 4.4%, no concordance 1.9%. For spherical subjective refraction, the rates were: high 65.3%, medium 20.8%, low 9.2%, no 4.6%. For cylindrical objective refraction concordance was: high 87.8%, medium 8.3%, low 2.1%, no 1.8%; cylindrical subjective refraction concordance was: high 81.5%, medium 12.2%, low 4.5%, no 1.7%.

The evaluation of Chronos binocular objective and subjective refraction versus standard methods demonstrated satisfactory concordance with standard refraction methods. This study highlights Chronos with the SightPilot algorithm as a dependable alternative for clinical refraction, streamlining patient assessments without compromising accuracy. Future research could assess the comfort levels experienced by both practitioners and patients, focusing on aspects such as duration and ergonomics.