Journal article
Deep learning network with differentiable dynamic programming for retina OCT surface segmentation
Biomedical optics express, Vol.14(7), pp.3190-3202
07/01/2023
DOI: 10.1364/BOE.492670
PMCID: PMC10368040
PMID: 37497505
Abstract
Multiple-surface segmentation in optical coherence tomography (OCT) images is a challenging problem, further complicated by the frequent presence of weak image boundaries. Recently, many deep learning-based methods have been developed for this task and yield remarkable performance. Unfortunately, due to the scarcity of training data in medical imaging, it is challenging for deep learning networks to learn the global structure of the target surfaces, including surface smoothness. To bridge this gap, this study proposes to seamlessly unify a U-Net for feature learning with a constrained differentiable dynamic programming module to achieve end-to-end learning for retina OCT surface segmentation to explicitly enforce surface smoothness. It effectively utilizes the feedback from the downstream model optimization module to guide feature learning, yielding better enforcement of global structures of the target surfaces. Experiments on Duke AMD (age-related macular degeneration) and JHU MS (multiple sclerosis) OCT data sets for retinal layer segmentation demonstrated that the proposed method was able to achieve subvoxel accuracy on both datasets, with the mean absolute surface distance (MASD) errors of 1.88 ± 1.96 μm and 2.75 ± 0.94 μm , respectively, over all the segmented surfaces.
Details
- Title: Subtitle
- Deep learning network with differentiable dynamic programming for retina OCT surface segmentation
- Creators
- Hui XieWeiyu XuYa Xing Wang - Capital UniversityXiaodong Wu
- Resource Type
- Journal article
- Publication Details
- Biomedical optics express, Vol.14(7), pp.3190-3202
- DOI
- 10.1364/BOE.492670
- PMID
- 37497505
- PMCID
- PMC10368040
- ISSN
- 2156-7085
- eISSN
- 2156-7085
- Grant note
- DOI: 10.13039/100000025, name: National Institute of Mental Health, award: 1U54HL165442; DOI: 10.13039/100000001, name: National Science Foundation, award: CCF-1733742, ECCS-2000425, ECCS-2133205
- Language
- English
- Date published
- 07/01/2023
- Academic Unit
- Electrical and Computer Engineering; Radiation Oncology; The Iowa Institute for Biomedical Imaging
- Record Identifier
- 9984444869502771
Metrics
6 Record Views