Skip to main content
main-content
Top

Hint

Swipe to navigate through the chapters of this book

2019 | OriginalPaper | Chapter

Learning to Segment Skin Lesions from Noisy Annotations

share
SHARE

Abstract

Deep convolutional neural networks have driven substantial advancements in the automatic understanding of images. Requiring a large collection of images and their associated annotations is one of the main bottlenecks limiting the adoption of deep networks. In the task of medical image segmentation, requiring pixel-level semantic annotations performed by human experts exacerbate this difficulty. This paper proposes a new framework to train a fully convolutional segmentation network from a large set of cheap unreliable annotations and a small set of expert-level clean annotations. We propose a spatially adaptive reweighting approach to treat clean and noisy pixel-level annotations commensurately in the loss function. We deploy a meta-learning approach to assign higher importance to pixels whose loss gradient direction is closer to those of clean data. Our experiments on training the network using segmentation ground truth corrupted with different levels of annotation noise show how spatial reweighting improves the robustness of deep networks to noisy annotations.
Literature
2.
go back to reference Goldberger, J., Ben-Reuven, E.: Training deep neural-networks using a noise adaptation layer. In: ICLR (2017) Goldberger, J., Ben-Reuven, E.: Training deep neural-networks using a noise adaptation layer. In: ICLR (2017)
3.
go back to reference Ibtehaz, N., Rahman, M. S.: Multiresunet: rethinking the U-Net architecture for multimodal biomedical image segmentation. arXiv preprint arXiv:​1902.​04049 (2019) Ibtehaz, N., Rahman, M. S.: Multiresunet: rethinking the U-Net architecture for multimodal biomedical image segmentation. arXiv preprint arXiv:​1902.​04049 (2019)
5.
go back to reference Long, et al.: Fully convolutional networks for semantic segmentation. In: IEEE CVPR, pp. 3431–3440 (2015) Long, et al.: Fully convolutional networks for semantic segmentation. In: IEEE CVPR, pp. 3431–3440 (2015)
7.
go back to reference Oktay, et al.: Attention U-Net: learning where to look for the pancreas. In: MIDL (2018) Oktay, et al.: Attention U-Net: learning where to look for the pancreas. In: MIDL (2018)
8.
go back to reference Ren, et al.: Learning to reweight examples for robust deep learning. In: ICML (2018) Ren, et al.: Learning to reweight examples for robust deep learning. In: ICML (2018)
10.
go back to reference Siegel, et al.: Cancer statistics. CA. Cancer J. Clin. 67(1), 7–30 (2017) Siegel, et al.: Cancer statistics. CA. Cancer J. Clin. 67(1), 7–30 (2017)
12.
go back to reference Vahdat, A.: Toward robustness against label noise in training deep discriminative neural networks. In: NIPS, pp. 5596–5605 (2017) Vahdat, A.: Toward robustness against label noise in training deep discriminative neural networks. In: NIPS, pp. 5596–5605 (2017)
13.
go back to reference Veit, et al.: Learning from noisy large-scale datasets with minimal supervision. In: IEEE CVPR, pp. 839–847 (2017) Veit, et al.: Learning from noisy large-scale datasets with minimal supervision. In: IEEE CVPR, pp. 839–847 (2017)
14.
go back to reference Venkatesh, et al.: A deep residual architecture for skin lesion segmentation. In: OR 2.0 Context-Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clinical Image-Based Procedures, and Skin Image Analysis, pp. 277–284 (2018) Venkatesh, et al.: A deep residual architecture for skin lesion segmentation. In: OR 2.0 Context-Aware Operating Theaters, Computer Assisted Robotic Endoscopy, Clinical Image-Based Procedures, and Skin Image Analysis, pp. 277–284 (2018)
15.
go back to reference Wang, et al.: Iterative learning with open-set noisy labels. In: IEEE CVPR, pp. 8688–8696 (2018) Wang, et al.: Iterative learning with open-set noisy labels. In: IEEE CVPR, pp. 8688–8696 (2018)
16.
go back to reference Xiao, et al.: Learning from massive noisy labeled data for image classification. In: IEEE CVPR, pp. 2691–2699 (2015) Xiao, et al.: Learning from massive noisy labeled data for image classification. In: IEEE CVPR, pp. 2691–2699 (2015)
17.
go back to reference Yuan, et al.: Automatic skin lesion segmentation using deep fully convolutional networks with jaccard distance. IEEE TMI 36(9), 1876–1886 (2017) Yuan, et al.: Automatic skin lesion segmentation using deep fully convolutional networks with jaccard distance. IEEE TMI 36(9), 1876–1886 (2017)
18.
go back to reference Zhang, et al.: Understanding deep learning requires rethinking generalization. In: ICLR (2017) Zhang, et al.: Understanding deep learning requires rethinking generalization. In: ICLR (2017)
Metadata
Title
Learning to Segment Skin Lesions from Noisy Annotations
Authors
Zahra Mirikharaji
Yiqi Yan
Ghassan Hamarneh
Copyright Year
2019
DOI
https://doi.org/10.1007/978-3-030-33391-1_24

Premium Partner