Macular holes are a common eye condition which result in visual impairment. We look at the application of deep convolutional neural networks to the problem of macular hole segmentation. We use the 3D U-Net architecture as a basis and experiment with a number of design variants. Manually annotating and measuring macular holes is time consuming and error prone, taking dozens of minutes to annotate a single 3D scan. Previous automated approaches to macular hole segmentation take minutes to segment a single 3D scan. We found that, in less than one second, deep learning models generate significantly more accurate segmentations than previous automated approaches (Jaccard index boost of 0.08 − 0.09) and expert agreement (Jaccard index boost of 0.13 − 0.20). We also demonstrate that an approach of architectural simplification, by greatly simplifying the network capacity and depth, results in a model which is competitive with state-of-the-art models such as residual 3D U-Nets.
Frawley, J., Willcocks, C. G., Habib, M., Geenen, C., Steel, D. H., Obara, B., …Mac Namee, B. (2021). Robust 3D U-Net Segmentation of Macular Holes.