This study assessed the anterior chamber depth (ACD) and iridocorneal angle using a portable smart eye camera (SEC) compared to the conventional slit-lamp microscope and anterior-segment optical coherence tomography (AS-OCT). This retrospective case-control study included 170 eyes from 85 Japanese patients. The correlation between the ACD evaluations conducted with the SEC and conventional slit-lamp was high (r = 0.814). The correlation between the Van-Herick Plus grade obtained using two devices was also high (r = 0.919). A high kappa value was observed for the Van-Herick Plus grading (Kappa = 0.757). A moderate correlation was observed between the ACD measured using AS-OCT and the slit-lamp image acquired with the conventional slit-lamp microscope and SEC (r = 0.609 and 0.641). A strong correlation was observed between the trabecular-iris angle (TIA) measured using AS-OCT and Van-Herick Plus grade obtained with the conventional slit-lamp microscope and SEC (r = 0.702 and 0.764). Strong correlations of ACD evaluation and high kappa value of the Van-Herick Plus grading indicated the adequate subjective assessment function of the SEC. Moderate correlations between the ACD objective measurement and evaluation and strong correlation between the TIA and Van-Herick Plus grade suggested the good objective assessment function of the SEC. The SEC demonstrated adequate performance for ACD evaluation and angle estimation.

The use of artificial intelligence (AI) in the diagnosis of dry eye disease (DED) remains limited due to the lack of standardized image formats and analysis models. To overcome these issues, we used the Smart Eye Camera (SEC), a video-recordable slit-lamp device, and collected videos of the anterior segment of the eye. This study aimed to evaluate the accuracy of the AI algorithm in estimating the tear film breakup time and apply this model for the diagnosis of DED according to the Asia Dry Eye Society (ADES) DED diagnostic criteria. Using the retrospectively corrected DED videos of 158 eyes from 79 patients, 22,172 frames were annotated by the DED specialist to label whether or not the frame had breakup. The AI algorithm was developed using the training dataset and machine learning. The DED criteria of the ADES was used to determine the diagnostic performance. The accuracy of tear film breakup time estimation was 0.789 (95% confidence interval (CI) 0.769–0.809), and the area under the receiver operating characteristic curve of this AI model was 0.877 (95% CI 0.861–0.893). The sensitivity and specificity of this AI model for the diagnosis of DED was 0.778 (95% CI 0.572–0.912) and 0.857 (95% CI 0.564–0.866), respectively. We successfully developed a novel AI-based diagnostic model for DED. Our diagnostic model has the potential to enable ophthalmology examination outside hospitals and clinics.

No abstract

No abstract

No abstract