Diagnosis of dry eye disease (DED) involves a combination of invasive, non-reproducible, and inaccurate tests. Artificial intelligence (AI) is a promising approach for improving the accuracy of DED diagnosis.1 However, its use remains limited due to non-standardized image formats and analysis models.2 An AI algorithm estimating the tear film breakup time has now been developed to diagnose DED using ocular surface videos.
The Asia Dry Eye Society (ADES) DED diagnostic criteria were used to develop this model, which used DED videos of 158 eyes from 79 patients. A total of 22,172 frames were annotated to label whether or not the frame had a breakup.2