If citing, it is advised that you check and use the publisher's definitive version for pagination, volume/issue, and date of publication details. And where the final published version is provided on the Research Portal, if citing you are again advised to check the publisher's website for any subsequent corrections.

The current study describes the development and assessment of innovative, machine learning (ML)-based approaches for automated detection and pixel-accurate measurements of regions with geographic atrophy (GA) in late-stage age-related macular degeneration (AMD) using optical coherence tomography systems. 900 OCT volumes, 100266 B-scans, and en face OCT images from 341 non-exudative AMD patients with or without GA were included in this study from both Cirrus (Zeiss) and Spectralis (Heidelberg) OCT systems. B-scan and en face level ground truth GA masks were created on OCT B-scan where the segmented ellipsoid zone (EZ) line, retinal pigment epithelium (RPE) line, and bruchs membrane (BM) line overlapped. Two deep learning-based approaches, B-scan level and en face level, were trained. The OCT B-scan model had detection accuracy of 91% and GA area measurement accuracy of 94%. The en face OCT model had detection accuracy of 82% and GA area measurement accuracy of 96% with primary target of hypertransmission on en face OCT. Accuracy was good for both devices tested (92–97%). Automated lesion size stratification for CAM cRORA definition of 250um minimum lesion size was feasible. High-performance models for automatic detection and segmentation of GA area were achieved using OCT systems and deep learning. The automatic measurements showed high correlation with the ground truth. The en face model excelled at identification of hypertransmission defects. The models performance generalized well across device types tested. Future development will include integration of both models to enhance feature detection across GA lesions as well as isolating hypertransmission defects without GA for pre-GA biomarker extraction.

Background: The development and testing of a deep learning (DL)-based approach for detection and measurement of regions of Ellipsoid Zone (EZ) At-Risk to study progression in nonexudative age-related macular degeneration (AMD). Methods: Used in DL model training and testing were 341 subjects with nonexudative AMD with or without geographic atrophy (GA). An independent dataset of 120 subjects were used for testing model performance for prediction of GA progression. Accuracy, specificity, sensitivity, and intraclass correlation coefficient (ICC) for DL-based EZ At-Risk percentage area measurement was calculated. Random forest-based feature ranking of EZ At-Risk was compared to previously validated quantitative OCT-based biomarkers. Results: The model achieved a detection accuracy of 99% (sensitivity = 99%; specificity = 100%) for EZ At-Risk. Automatic EZ At-Risk measurement achieved an accuracy of 90% (sensitivity = 90%; specificity = 84%) and the ICC compared to ground truth was high (0.83). In the independent dataset, higher baseline mean EZ At-Risk correlated with higher progression to GA at year 5 (p < 0.001). EZ At-Risk was a top ranked feature in the random forest assessment for GA prediction. Conclusions: This report describes a novel high performance DL-based model for the detection and measurement of EZ At-Risk. This biomarker showed promising results in predicting progression in nonexudative AMD patients.