Robust 3D U-Net Segmentation of Macular Holes

Frawley, Jonathan; Willcocks, Chris G.; Habib, Maged; Geenen, Caspar; Steel, David H.; Obara, Boguslaw; Pakrashi, Arjun; Rushe, Ellen; Bazargani, Mehran Hossein Zadeh; Mac Namee, Brian

Abstract

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.

Citation

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.

• Conference Name: The 29th Irish Conference on Artificial Intelligence and Cognitive Science 2021, Dublin, Republic of Ireland, December 9-10, 2021
• Conference Location: Dublin, Ireland
• Start Date: Dec 9, 2021
• End Date: Dec 10, 2021
• Acceptance Date: Nov 22, 2021
• Online Publication Date: Dec 8, 2021
• Publication Date: 2021
• Deposit Date: Oct 23, 2022
• Publicly Available Date: Oct 24, 2022
• Volume: 3105
• Pages: 36-47
• Series Title: CEUR Workshop Proceedings
• Publisher URL: http://ceur-ws.org/Vol-3105/