Synergistic Image and Feature Adaptation: Towards Cross-Modality Domain Adaptation for Medical Image Segmentation

By Cheng Chen, Qi-Yan Dou, Hao Chen, Jing Qin, P. Heng. AAAI Conference on Artificial Intelligence, 2019

374 citations recorded by the source corpus.

This paper presents a novel unsupervised domain adaptation framework, called Synergistic Image and Feature Adaptation (SIFA), to effectively tackle the problem of domain shift. Domain adaptation has become an important and hot topic in recent studies on deep learning, aiming to recover performance degradation when applying the neural networks to new testing domains. Our proposed SIFA is an elegant learning diagram which presents synergistic fusion of adaptations from both image and feature perspectives. In particular, we simultaneously transform the appearance of images across domains and enhance domain-invariance of the extracted features towards the segmentation task. The feature encoder layers are shared by both perspectives to grasp their mutual benefits during the end-to-end learning procedure. Without using any annotation from the target domain, the learning of our unified model is guided by adversarial losses, with multiple discriminators employed from various aspects. We have extensively validated our method with a challenging application of crossmodality medical image segmentation of cardiac structures. Experimental results demonstrate that our SIFA model recovers the degraded performance from 17.2% to 73.0%, and outperforms the state-of-the-art methods by a significant margin.

Subfields: Artificial Intelligence, Biomedical Engineering, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/0597cf497bdd03fa2d4b0cec40036dc1faf4de62

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. P. Heng
2. Qi-Yan Dou
3. Hao Chen
4. Jing Qin
5. Lequan Yu
6. Cheng Chen
7. Garrett Wilson
8. Diane Joyce Cook

In Biomedical Engineering

1. P. Heng
2. Qi-Yan Dou
3. Lequan Yu
4. Hao Chen
5. Diane Joyce Cook
6. Jing Qin
7. James S. Duncan
8. Garrett Wilson

In Radiology, Nuclear Medicine and Imaging

1. P. Heng
2. Qi-Yan Dou
3. Hao Chen
4. Jing Qin
5. Lequan Yu
6. Diane Joyce Cook
7. Garrett Wilson
8. Cheng Chen

Citations in the corpus

• Multi-site fMRI Analysis Using Privacy-preserving Federated Learning and Domain Adaptation: ABIDE Results (2020, background): “, 2018) have shown promising results of utilizing domain adaptation techniques to assist heterogeneous data analysis, including the applications in medical image analysis areas (Chen et al., 2019; Yang et al., 2019a).”
• Medical Image Segmentation Using Deep Learning: A Survey (2020, neutral): “[112] also adopted a similar”
• Learning from Extrinsic and Intrinsic Supervisions for Domain Generalization (2020, background): “The general idea of domain adaption is to align the source domain and target domain distributions in the input level [3,19], semantic feature level [9,34], or output space level [5,39,40,43,45].”
• Unsupervised Bidirectional Cross-Modality Adaptation via Deeply Synergistic Image and Feature Alignment for Medical Image Segmentation (2020, methodology): “The higher discriminator loss of our SIFA model than [20] indicates that the feature distributions have been aligned better such that it becomes harder for the discriminator to classify the domain of the segmentation predictions.”
• A Survey of Unsupervised Deep Domain Adaptation (2018, methodology): “ormation is available from 28 Garrett Wilson and Diane J. Cook a simulator [116], in a medical context for chest X-ray segmentation [28], 3D CT scans to X-ray segmentation[238],MRItoCTscansegmentation[29],andMRIsegmentation[161],inlowresource situations (where there are very few target data points) ”
• SynthSeg: Segmentation of brain MRI scans of any contrast and resolution without retraining (2021, neutral): “SIFA [11] 8 : For this method, we added an online data augmentation step during training, where we apply the same spatial, intensity and bias augmentations as for SynthSeg (Dice improvement of 0.18 on the validation set).”
• MetaCorrection: Domain-aware Meta Loss Correction for Unsupervised Domain Adaptation in Semantic Segmentation (2021, neutral): “The general idea of UDA semantic segmentation methods is to perform domain alignment through adversarial learning [2, 6, 7, 9, 11, 20, 22, 26, 40, 41] or utilize self-training strategy on target samples to adapt the segmentation models [4, 17, 23, 28, 30, 35, 46, 48, 49, 50, 51].”
• Advancing Medical Imaging Informatics by Deep Learning-Based Domain Adaptation (2020, background): “For example, MRI achieves higher resolution for soft tissue imaging compared to CT [13].”