Liefers B, Taylor P, Alsaedi A, et al. - In the present study, the researchers developed and validated a deep learning model for segmentation of 13 features correlated with neovascular and atrophic age-related macular degeneration. Data for model development were collected from 307 optical coherence tomography volumes. With this data, a deep neural network was trained to perform voxel-level segmentation of the 13 most common abnormalities (features). The standard of automatic segmentation is consistent with that of experienced graders for most features, which surpass human performance for some features. In the current clinical routine, the quantified parameters given by the model can be used and open possibilities for more treatment response testing beyond clinical trials can be used.