Semi-supervised Medical Image Segmentation via Learning Consistency Under Transformations
Open accessThe DOI matches its Crossref record; 16 of the 16 references it cites were found in a registry or an index.
Abstract
The scarcity of labeled data often limits the application of supervised deep learning techniques for medical image segmentation. This has motivated the development of semi-supervised techniques that learn from a mixture of labeled and unlabeled images. In this paper, we propose a novel semi-supervised method that, in addition to supervised learning on labeled training images, learns to predict segmentations consistent under a given class of transformations on both labeled and unlabeled images. More specifically, in this work we explore learning equivariance to elastic deformations. We implement this through: (1) a Siamese architecture with two identical branches, each of which receives a differently transformed image, and (2) a composite loss function with a supervised segmentation loss term and an unsupervised term that encourages segmentation consistency between the predictions of the two branches. We evaluate the method on a public dataset of chest radiographs with segmentations of anatomical structures using 5-fold cross-validation. The proposed method reaches significantly higher segmentation accuracy compared to supervised learning. This is due to learning transformation consistency on both labeled and unlabeled images, with the latter contributing the most. We achieve the performance comparable to state-of-the-art chest X-ray segmentation methods while using substantially fewer labeled images.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
U-Net: Convolutional Networks for Biomedical Image Segmentation
Ronneberger, O., Fischer, P., & Brox, T. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24574-4_28 · Cited by 99,404
Shiraishi, J., Katsuragawa, S., Ikezoe, J., Matsumoto, T., Kobayashi, T., Komatsu, K.-I., Matsui, M., Fujita, H., Kodera, Y., & Doi, K. (2000) · American Journal of Roentgenology · doi:10.2214/ajr.174.1.1740071 · Cited by 1,004 · cited 2×
“rse (tout) 1 can be computed. Semi-Supervised Segmentation via Transformation Consistency 5 3 Experiments Dataset and validation We used the Japanese Society of Radiological Technology (JSRT) dataset [15], which contains 247 posterior-anterior chest radiographs with a resolution 2048 2048, 0:175mm pixel size and 12-bi…”
Methodvan Ginneken, B., Stegmann, M. B., & Loog, M. (2006) · Medical Image Analysis · doi:10.1016/j.media.2005.02.002 · Cited by 539 · cited 5×
“Seg-mentations of the left and right lung fields, left and right clavicles and the heart were made available by [16].”
InfluentialRegularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning
Sajjadi, M. S. M., Javanmardi, M., & Taşdizen, T. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1606.04586 · Cited by 532 · cited 1×
“This idea was already explored before and applied to the prediction of image-level labels: image classification [13,10] and landmark coordinate regression [6].”
MethodSemi-Supervised Learning with Ladder Networks
Rasmus, A., Harri Valpola, B., Honkala, M., Berglund, M., & Raiko, T. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1507.02672 · Cited by 527 · cited 1×
“This idea was already explored before and applied to the prediction of image-level labels: image classification [13,10] and landmark coordinate regression [6].”
MethodFully Convolutional Architectures for Multiclass Segmentation in Chest Radiographs
Novikov, A. A., Lenis, D., Major, D., Hladuvka, J., Wimmer, M., & Buhler, K. (2018) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2018.2806086 · Cited by 252
SCAN: Structure Correcting Adversarial Network for Organ Segmentation in Chest X-Rays
Dai, W., Dong, N., Wang, Z., Liang, X., Zhang, H., & Xing, E. P. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-00889-5_30 · Cited by 193 · cited 3×
“[3] and Novikov et al.”
MethodImproving Landmark Localization with Semi-Supervised Learning
Honari, S., Molchanov, P., Tyree, S., Vincent, P., Pal, C., & Kautz, J. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00167 · Cited by 172
Semi-supervised Skin Lesion Segmentation via Transformation Consistent Self-ensembling ModelPreprint
Li, X., Yu, L., Chen, H., Fu, C., & Heng, P. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1808.03887 · Cited by 146 · cited 3×
“to labeled examples according to prior adjacency information. Self-ensembling approaches learn from synthetic labels on unlabeled data, which are constructed using past iterations of the same network [7,9]. In this paper, we propose an approach in which we learn consistency under transformations on both labeled and un…”
BackgroundMethodInfluentialExploiting the potential of unlabeled endoscopic video data with self-supervised learning
Ross, T., Zimmerer, D., Vemuri, A., Isensee, F., Wiesenfarth, M., Bodenstedt, S., Both, F., Kessler, P., Wagner, M., Müller, B., et al. (2018) · International Journal of Computer Assisted Radiology and Surgery · doi:10.1007/s11548-018-1772-0 · Cited by 135 · cited 1×
“lysis through utilizing unlabeled data [2]. Selfsupervised approaches learn to perform an auxiliary task related to the target prediction task on unlabeled data, examples including image colorization [12] and predicting one image modality from another [5]. Auxiliary manifold embedding [1] learns to place unlabeled exa…”
Semi-supervised Deep Learning for Fully Convolutional Networks
Baur, C., Albarqouni, S., & Navab, N. (2017) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-66179-7_36 · Cited by 118 · cited 1×
“Auxiliary manifold embedding [1] learns to place unlabeled examples far or close in the feature space with respect to each other and to labeled examples according to prior adjacency information.”
BackgroundCheplygina, V., de Bruijne, M., & Pluim, J. P. W. (2019) · Medical Image Analysis · doi:10.1016/j.media.2019.03.009 · Cited by 93
Frid-Adar, M., Ben-Cohen, A., Amer, R., & Greenspan, H. K. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-00946-5_17 · Cited by 82 · cited 5×
“We compare the proposed method trained in the small data (20 labeled images) and full supervision regimes with state-of-the-art methods [14,3,8,4] and the inter-observer agreement [16].”
MethodInfluentialDeep Semi-supervised Segmentation with Weight-Averaged Consistency Targets
Perone, C. S., & Cohen‐Adad, J. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-00889-5_2 · Cited by 72 · cited 2×
“Self-ensembling approaches [7,9] are similar to the proposed approach in that they use a transformation consistency prior: they construct their synthetic labels on unlabeled data using predictions on differently transformed inputs.”
BackgroundMethodMS-Net: Mixed-Supervision Fully-Convolutional Networks for Full-Resolution Segmentation
Shah, M., Merchant, S. N., & Awate, S. P. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-00937-3_44 · Cited by 59 · cited 4×
“The proposed method substantially outperformed MS-Net [14], the only weakly supervised method evaluated on JSRT that is known to us.”
BackgroundMethodInfluentialRetinal Image Understanding Emerges from Self-Supervised Multimodal Reconstruction
Hervella, Á. S., Rouco, J., Novo, J., & Ortega, M. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-00928-1_37 · Cited by 43 · cited 1×
“Selfsupervised approaches learn to perform an auxiliary task related to the target prediction task on unlabeled data, examples including image colorization [12] and predicting one image modality from another [5].”
Background
Cited by
Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation
Tajbakhsh, N., Jeyaseelan, L., Li, Q., Chiang, J. N., Wu, Z., & Ding, X. (2020) · Medical Image Analysis · doi:10.1016/j.media.2020.101693 · Cited by 1,095
Deep learning for chest X-ray analysis: A survey
Çallı, E., Sogancioglu, E., van Ginneken, B., van Leeuwen, K. G., & Murphy, K. (2021) · Medical Image Analysis · doi:10.1016/j.media.2021.102125 · Cited by 506
Shape-Aware Semi-supervised 3D Semantic Segmentation for Medical Images
Li, S., Zhang, C., & He, X. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-59710-8_54 · Cited by 426
You, C., Zhou, Y., Zhao, R., Staib, L., & Duncan, J. S. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2022.3161829 · Cited by 332
Luo, X., Liao, W., Chen, J., Song, T., Chen, Y., Zhang, S., Chen, N., Wang, G., & Zhang, S. (2021) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-87196-3_30 · Cited by 270
Metadata
- Authors
- Bortsova, G., Dubost, F., Hogeweg, L., Katramados, I., & de Bruijne, M.
- Venue
- Lecture Notes in Computer Science
- Publisher
- Springer International Publishing
- Published
- 2019
- Pages
- 810-818
- ISSN
- 0302-9743 (ISSN-L)
- 1611-3349
- OpenAlex
- W2979433110
- Cited by
- 183
- References
- 16
- FWCI
- 6.33
- Type
- conference paper
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.