Deep Adversarial Networks for Biomedical Image Segmentation Utilizing Unannotated Images

By Yizhe Zhang, Lin F. Yang, Jianxu Chen, Maridel Fredericksen, David P. Hughes, Danny Ziyi Chen. 2017

436 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

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

In Computer Vision and Pattern Recognition

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Visual Feature Attribution Using Wasserstein GANs (2017, methodology): “In the context of medical images, GANs have been applied to super-resolution in retinal fundus images [34], for semi-supervised cardiac segmentation [64], synthesising computed tomography images from MR images[40, 60] and intraoperative motion modelling [23].”
Generative Adversarial Network in Medical Imaging: A Review (2018, background): “Zhang et al. (2017c) proposed to use both annotated and unannotated images in the segmentation pipeline.”
Transformation-Consistent Self-Ensembling Model for Semisupervised Medical Image Segmentation (2019, methodology): “that our method gains higher performance consistently than other methods [18, 28, 52, 59] in both 10% and 20% settings, respectively.”
Semi-supervised Medical Image Segmentation through Dual-task Consistency (2020, methodology): “Comparison with Other Semi-supervised Methods: We compared our framework with six state-of-the-art semisupervised segmentation methods, including deep adversarial network (DAN) (Zhang et al. 2017), entropy minimization approach (Entropy Mini) (Vu et al.”
Learning with Limited Annotations: A Survey on Deep Semi-Supervised Learning for Medical Image Segmentation (2022, background): “methods always contain a discriminator to distinguish the inputs from labeled annotations or unlabeled predictions [61], [93], [94], [95].”
Unsupervised Reverse Domain Adaptation for Synthetic Medical Images via Adversarial Training (2017, positive): “ramework. Adversarial Training for Biomedical Imaging: Various kinds of adversarial training has recently been used for a variety of medical imaging tasks including noise reduction [33], segmentation [35,16], detection [11], reconstruction [15], classiﬁcation [34] and image synthesis [18]. Osokin et”
GANs for Medical Image Analysis (2018, background): “[73] proposes DAN — a combination of DCAN [74] and VGG16 — which is trained with both supervised and unsupervised strategies to provide high-quality segmentation masks for unseen images.”
SimCVD: Simple Contrastive Voxel-Wise Representation Distillation for Semi-Supervised Medical Image Segmentation (2021, background): “Zhang et al.[15] proposed a novel deep adversarial network termed DAN by introducing adversarial loss to compute the similarity between the predicted segmentation maps of labeled and unlabeled data.”
A Review of Deep-Learning-Based Medical Image Segmentation Methods (2021, background): “[59] proposed a new deep adversarial network (DAN) for medical image segmentation, with the goal of obtaining good segmentation results on both annotated and unannotated images.”
Factorised spatial representation learning: application in semi-supervised myocardial segmentation (2018, methodology): “Also we do not require any pairs of image and masks for discriminator training as in [20], and we retain all information, in contrast to [3] which preserves only task relevant information.”
Enabling a Single Deep Learning Model for Accurate Gland Instance Segmentation: A Shape-Aware Adversarial Learning Framework (2020, background): “Zhang et al. [17] proposed a generative deep learning framework to make use of additional unannotated images.”
SynthSeg: Segmentation of brain MRI scans of any contrast and resolution without retraining (2021, positive): “While such transforms often relies on handcrafted (and thus suboptimal) parameters, recent semi-supervised methods, such as adversarial augmentation, explicitly optimise the augmentation parameters during training [73, 9, 12].”
Medical Image Synthesis for Data Augmentation and Anonymization using Generative Adversarial Networks (2018, neutral): “Segmentation using GANs was demonstrated in [22,21].”
Semi-Supervised Medical Image Segmentation via Cross Teaching between CNN and Transformer (2021, methodology): “, 2019), Deep Adversarial Network (DAN) (Zhang et al., 2017), Deep Co-Training (DCT) (Qiao et al.”
Synthetic Data for Deep Learning (2019, background): “A related but different field considers unsupervised approaches to segmentation and other computer vision problems based on adversarial architectures, including learning to segment via cut-and-paste [487], unsupervised segmentation by moving objects between pairs of images with inpainting [434], seg”