Predicting incremental and future visual change in neovascular age-related macular degeneration using deep learning
Fu, Dun Jack; Keane, Pearse (2021), Predicting incremental and future visual change in neovascular age-related macular degeneration using deep learning, Dryad, Dataset, https://doi.org/10.5061/dryad.573n5tb5d
Purpose To evaluate the predictive utility of quantitative imaging biomarkers, acquired automatically from optical coherence tomography (OCT) scans, of cross-sectional and future visual outcomes of patients with neovascular age-related macular degeneration (AMD) starting anti-vascular endothelial growth factor (VEGF) therapy.
Design Retrospective cohort study.
Methods Automatic segmentation was carried out by applying a deep learning segmentation algorithm to 137,379 OCT scans from 6467 eyes of 3261 patients with neovascular AMD between 2007 and 2017 at Moorfields Eye Hospital (a large, UK single-centre). Treatment-naïve, first-treated eyes were taken forward for analysis - 926 eyes of 926 patients. Main outcome measures were correlation coefficients (R2) and mean absolute error (MAE) between quantitative OCT (qOCT) parameters and cross-sectional visual-function; VA at distant timepoints (up to 12 months post-baseline); the incremental VA-change from an individual injection.
Results VA at distant timepoints could be predicted: R2 0.79 (MAE 5.0 ETDRS letters) and R2 0.63 (MAE 7.2) post-injection 3 and at 12 months post-baseline (both p < 0.001), respectively. Best performing models included both baseline qOCT parameters and treatment-response. Furthermore, we present proof-of-principle evidence that the incremental change in VA from an injection can be predicted: R2 0.13 (MAE 5.6) for injection 2 and R2 0.07 (MAE 5.0) for injection 3 (both p < 0.01).
Conclusions Automatic segmentation enables rapid acquisition of quantitative and reproducible OCT biomarkers with potential to inform treatment decisions in the care of neovascular AMD. This furthers development of point-of-care decision-aid systems for personalized medicine.
UK Research and Innovation