CATALYZE: a deep learning approach for cataract assessment and grading on SS-OCT images

Purpose: To assess a new objective deep learning model cataract grading method based on swept-source optical coherence tomography (SS-OCT) scans provided by the Anterion. Setting: Single-center study at the Rothschild Foundation, Paris, France. Design: Prospective cross-sectional study. Methods: All patients consulting for cataract evaluation and consenting to study participation were included. History of previous ocular surgery, cornea or retina disorders, and ocular dryness were exclusion criteria. Our CATALYZE pipeline was applied to Anterion image providing layerwise cataract metrics and an overall clinical significance index (CSI) of cataract. Ocular scatter index (OSI) was also measured with a double-pass aberrometer (OQAS) and compared with our CSI. Results: 548 eyes were included, 331 in the development set (48 with cataract and 283 controls) and 217 in the validation set (85 with cataract and 132 controls) of 315 patients aged 19 to 85 years (mean ± SD: 50 ± 21 years). The CSI correlated with the OSI ( r 2 = 0.87, P < .01). CSI area under the receiver operating characteristic curve (AUROC) was comparable with OSI AUROC (0.985 vs 0.981 respectively, P > .05) with 95% sensitivity and 95% specificity. Conclusions: The deep learning pipeline CATALYZE based on Anterion SS-OCT may be a reliable and comprehensive objective cataract grading method.