Related Experiment Video
Updated: Feb 28, 2026
05:10
Author Spotlight: An Automated Method for Assessing Visual Acuity in Infants and Toddlers Using an Eye-Tracking System
Published on: March 17, 2023
3.8K
Development and Clinical Validation of an Artificial Intelligence-Based Automated Visual Acuity Testing System
Kelvin Zhenghao Li1,2, Hnin Hnin Oo1, Kenneth Chee Wei Liang1
1Department of Ophthalmology, National Healthcare Group Eye Institute, Tan Tock Seng Hospital, Singapore 308433, Singapore.
Life (Basel, Switzerland)
|February 27, 2026
Summary
This study presents an AI-powered automated visual acuity (VA) testing system that accurately assesses vision using speech and image recognition. The system demonstrates feasibility for self-administered, clinic-based eye exams.
Area of Science:
- Ophthalmology
- Artificial Intelligence
- Medical Technology
Background:
- Automated visual acuity (VA) testing systems can enhance clinic efficiency.
- Integrating artificial intelligence (AI) offers potential for self-administered eye assessments.
- Current VA testing methods may require significant clinician time and resources.
Purpose of the Study:
- To develop and validate an AI-driven automated visual acuity (VA) testing system.
- To enable self-administered, clinic-based VA assessment using speech and image recognition.
- To evaluate the accuracy, reliability, and user experience of the automated system.
Main Methods:
- Developed an AI system using Whisper speech recognition and pose estimation.
- Incorporated a state-driven interface for guided, sequential testing.
- Validated laboratory performance and compared automated vs. manual VA testing in a clinical setting.
Main Results:
- AI model significantly reduced word error rates for letter and number recognition.
- Pose detection accurately identified occluder use.
- Automated unaided VA showed good agreement with manual testing (ICC=0.77); pinhole VA showed moderate agreement (ICC=0.63).
Conclusions:
- The AI-based automated VA system is accurate, reliable, and user-friendly.
- The system demonstrates feasibility for clinical implementation.
- Automated VA testing shows promise for efficient and accessible vision assessment.
