Unsupervised Domain Adaptation from Axial to Short-Axis Multi-Slice Cardiac MR Images by Incorporating Pretrained Task Networks
Not on the map. Why?
Abstract (arXiv)
Anisotropic multi-slice Cardiac Magnetic Resonance (CMR) Images are conventionally acquired in patient-specific short-axis (SAX) orientation. In specific cardiovascular diseases that affect right ventricular (RV) morphology, acquisitions in standard axial (AX) orientation are preferred by some investigators, due to potential superiority in RV volume measurement for treatment planning. Unfortunately, due to the rare occurrence of these diseases, data in this domain is scarce. Recent research in deep learning-based methods mainly focused on SAX CMR images and they had proven to be very successful. In this work, we show that there is a considerable domain shift between AX and SAX images, and therefore, direct application of existing models yield sub-optimal results on AX samples. We propose a novel unsupervised domain adaptation approach, which uses task-related probabilities in an attention mechanism. Beyond that, cycle consistency is imposed on the learned patient-individual 3D rigid transformation to improve stability when automatically re-sampling the AX images to SAX orientations. The network was trained on 122 registered 3D AX-SAX CMR volume pairs from a multi-centric patient cohort. A mean 3D Dice of 0.86±0.06 for the left ventricle, 0.65±0.08 for the myocardium, and 0.77±0.10 for the right ventricle could be achieved. This is an improvement of 25% in Dice for RV in comparison to direct application on axial slices. To conclude, our pre-trained task module has neither seen CMR images nor labels from the target domain, but is able to segment them after the domain gap is reduced. Code: https://github.com/Cardio-AI/3d-mri-domain-adaptation
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 30 of its 45 references that are in the map. First posted on arXiv 19 Jan 2021.
Builds on
30 of 45 references.
References · 30
- 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
- 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
- 2015Spatial Transformer Networkscited by 1,423
- 2017CyCADA: Cycle-Consistent Adversarial Domain Adaptationcited by 1,079
- 2017Maximum Classifier Discrepancy for Unsupervised Domain Adaptationcited by 682
- 2018Deep Visual Domain Adaptation: A Surveycited by 634
- 2018Albumentations: fast and flexible image augmentationscited by 558
- 2019Confidence Regularized Self-Trainingcited by 418
- 2018A Survey of Unsupervised Deep Domain Adaptationcited by 361
- 2019Bidirectional Learning for Domain Adaptation of Semantic Segmentationcited by 308
- 2019Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentationcited by 258
- 2015BoxSup: Exploiting Bounding Boxes to Supervise Convolutional Networks for Semantic Segmentationcited by 238
18 more
- 2016Seed, Expand and Constrain: Three Principles for Weakly-Supervised Image Segmentationcited by 198
- 2015Weakly- and Semi-Supervised Learning of a DCNN for Semantic Image Segmentationcited by 189
- 2019Unsupervised Learning of Probabilistic Diffeomorphic Registration for Images and Surfacescited by 186
- 2018An Unsupervised Learning Model for Deformable Medical Image Registrationcited by 150
- 2016The Importance of Skip Connections in Biomedical Image Segmentationcited by 126
- 2017Anatomically Constrained Neural Networks (ACNN): Application to Cardiac Image Enhancement and Segmentationcited by 109
- 2020CrDoCo: Pixel-level Domain Transfer with Cross-Domain Consistencycited by 107
- 2017Adversarial Dropout Regularizationcited by 106
- 2017ROAD: Reality Oriented Adaptation for Semantic Segmentation of Urban Scenescited by 95
- 2020Unsupervised Domain Adaptation in Semantic Segmentation: a Reviewcited by 88
- 2018SPIGAN: Privileged Adversarial Learning from Simulationcited by 47
- 2019Curriculum Model Adaptation with Synthetic and Real Data for Semantic Foggy Scene Understandingcited by 40
- 2019Anatomical Priors in Convolutional Networks for Unsupervised Biomedical Segmentationcited by 40
- 2017An Exploration of 2D and 3D Deep Learning Techniques for Cardiac MR Image Segmentationcited by 36
- 2019Unsupervised Deformable Image Registration Using Cycle-Consistent CNNcited by 22
- 2019Adversarial Learning and Self-Teaching Techniques for Domain Adaptation in Semantic Segmentationcited by 22
- 2020Unsupervised Domain Adaptation with Multiple Domain Discriminators and Adaptive Self-Trainingcited by 10
- 2020How well do U-Net-based segmentation trained on adult cardiac magnetic resonance imaging data generalise to rare congenital heart diseases for surgical planning?cited by 2
Built on by
Citing papers per year: 2022: 2.
2 papers.