1,285 Hits in 5.4 sec

Enhancing MR Image Segmentation with Realistic Adversarial Data Augmentation [article]

Chen Chen, Chen Qin, Cheng Ouyang, Zeju Li, Shuo Wang, Huaqi Qiu, Liang Chen, Giacomo Tarroni, Wenjia Bai, Daniel Rueckert
2022 arXiv   pre-print
We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data.  ...  AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training  ...  In this work, four different image transformation models are employed to resemble realistic imaging variations in MR imaging.  ... 
arXiv:2108.03429v2 fatcat:m24wykdkbna3fdtq2t5qdlgq2i

Realistic Adversarial Data Augmentation for MR Image Segmentation [article]

Chen Chen, Chen Qin, Huaqi Qiu, Cheng Ouyang, Shuo Wang, Liang Chen, Giacomo Tarroni, Wenjia Bai, Daniel Rueckert
2020 arXiv   pre-print
In this work, we propose an adversarial data augmentation method for training neural networks for medical image segmentation.  ...  MR imaging: bias field.  ...  Discussion and Conclusion In this work, we presented a realistic adversarial data augmentation method to improve the generalization and robustness for neural network-based medical image segmentation.  ... 
arXiv:2006.13322v1 fatcat:3kbkfjurajgpjeqgl57ki5b24m

Learning Data Augmentation for Brain Tumor Segmentation with Coarse-to-Fine Generative Adversarial Networks [article]

Tony C.W Mok, Albert C.S Chung
2018 arXiv   pre-print
In this paper, we propose a novel automatic data augmentation method that uses generative adversarial networks to learn augmentations that enable machine learning based method to learn the available annotated  ...  While it is often easy for researchers to use data augmentation to expand the size of training sets, constructing and generating generic augmented data that is able to teach the network the desired invariance  ...  from generating a set of realistic MR images with the identical shape and context of the brain.  ... 
arXiv:1805.11291v2 fatcat:ikvianu6jzduxezi5yubsanyym

Enhanced Magnetic Resonance Image Synthesis with Contrast-Aware Generative Adversarial Networks [article]

Jonas Denck, Jens Guehring, Andreas Maier, Eva Rothgang
2021 arXiv   pre-print
With the rise of generative deep learning models, approaches for the synthesis of MR images are developed to either synthesize additional MR contrasts, generate synthetic data, or augment existing data  ...  This approach enables us to synthesize MR images with adjustable image contrast.  ...  of realistic MR images with the intended MR contrast.  ... 
arXiv:2102.09386v2 fatcat:o3wj6arm75ds7ef5vulfwsifsu

Combining Shape Priors with Conditional Adversarial Networks for Improved Scapula Segmentation in MR images [article]

Arnaud Boutillon, Bhushan Borotikar, Valérie Burdin, Pierre-Henri Conze
2020 arXiv   pre-print
This paper proposes an automatic method for scapula bone segmentation from Magnetic Resonance (MR) images using deep learning.  ...  the model by promoting realistic delineations.  ...  All networks were trained using data augmentation since the amount of available training data was limited.  ... 
arXiv:1910.08963v3 fatcat:g2jxwjypv5a45bk7p7l3qaisa4

Infinite Brain MR Images: PGGAN-based Data Augmentation for Tumor Detection [article]

Changhee Han, Leonardo Rundo, Ryosuke Araki, Yujiro Furukawa, Giancarlo Mauri, Hideki Nakayama, Hideaki Hayashi
2019 arXiv   pre-print
To fill the data lack in the real image distribution, we synthesize brain contrast-enhanced Magnetic Resonance (MR) images---realistic but completely different from the original ones---using Generative  ...  Due to the lack of available annotated medical images, accurate computer-assisted diagnosis requires intensive Data Augmentation (DA) techniques, such as geometric/intensity transformations of original  ...  , as we can discard failed cases for better data augmentation.  ... 
arXiv:1903.12564v1 fatcat:tdeyjwat2fduraxrseb5hzrsd4

Deep Learning-Based Forearm Subcutaneous Veins Segmentation

Zaineb Shah, Syed Ayaz Ali Shah, Aamir Shahzad, Ahmad Fayyaz, Shoaib Khaliq, Ali Zahir, Goh Chuan Meng
2022 IEEE Access  
INDEX TERMS Forearm subcutaneous veins, generative adversarial networks, image segmentation, medical image analysis.  ...  These networks generate realistic results by learning data mapping from one state to another.  ...  The sample images for data augmentation along with ground truth images are shown in Figure 1 . B.  ... 
doi:10.1109/access.2022.3167691 fatcat:psmdoj3vjnfn7o37cxhur2szze

Enhanced Magnetic Resonance Image Synthesis with Contrast-Aware Generative Adversarial Networks

Jonas Denck, Jens Guehring, Andreas Maier, Eva Rothgang
2021 Journal of Imaging  
With the rise of generative deep learning models, approaches for the synthesis of MR images are developed to either synthesize additional MR contrasts, generate synthetic data, or augment existing data  ...  Therefore, we trained a generative adversarial network (GAN) with a separate auxiliary classifier (AC) network to generate synthetic MR knee images conditioned on various acquisition parameters (repetition  ...  Moreover, GANs were used for enhanced image denoising [26] for brain MR images and the generation of additional training data for brain tissue segmentation networks [27] .  ... 
doi:10.3390/jimaging7080133 pmid:34460769 pmcid:PMC8404922 fatcat:gka5yg3trbasrbhpc3gt2lkcmq

Adversarial Image Synthesis for Unpaired Multi-modal Cardiac Data [chapter]

Agisilaos Chartsias, Thomas Joyce, Rohan Dharmakumar, Sotirios A. Tsaftaris
2017 Lecture Notes in Computer Science  
We demonstrate the application of this approach in synthesising cardiac MR images from CT images, using a dataset of MR and CT images coming from different patients.  ...  Since there can be no direct evaluation of the synthetic images, as no ground truth images exist, we demonstrate their utility by leveraging our synthetic data to achieve improved results in segmentation  ...  mask into realistic MR images and corresponding segmentation masks.  ... 
doi:10.1007/978-3-319-68127-6_1 fatcat:tidos465frdkrglrbziddxbwbu

Glioblastoma Synthesis and Segmentation with 3D Multi-Modal MRI: A Study using Generative Adversarial Networks

Edmond Wang
2021 International Journal on Computational Science & Applications  
Nonetheless, accurate segmentation and realistic image synthesis remain challenging tasks.  ...  The use of Deep Learning - in particular CNNs and GANs - have become prominent in dealing with various image segmentation and detection tasks.  ...  As a result, generative adversarial networks (GANs) have come to be used for image synthesis and data augmentation.  ... 
doi:10.5121/ijcsa.2021.11601 fatcat:4wjwcix2jrdrlckuoiwxnelhqq

High-resolution medical image synthesis using progressively grown generative adversarial networks [article]

Andrew Beers, James Brown, Ken Chang, J. Peter Campbell, Susan Ostmo, Michael F. Chiang, Jayashree Kalpathy-Cramer
2018 arXiv   pre-print
We also show that fine-grained details associated with pathology, such as retinal vessels or tumor heterogeneity, can be preserved and enhanced by including segmentation maps as additional channels.  ...  Generative adversarial networks (GANs) are a class of unsupervised machine learning algorithms that can produce realistic images from randomly-sampled vectors in a multi-dimensional space.  ...  We present an application of the PGGAN method to retinal fundus and MRI data, specifically for the diseases of retinopathy of prematurity and glioma.  ... 
arXiv:1805.03144v2 fatcat:zcj53gvqkbgidfonkx5hpjrqrm

Capturing Variabilities from Computed Tomography Images with Generative Adversarial Networks [article]

Umair Javaid, John A. Lee
2018 arXiv   pre-print
With the advent of Deep Learning (DL) techniques, especially Generative Adversarial Networks (GANs), data augmentation and generation are quickly evolving domains that have raised much interest recently  ...  To deal with this limitation, different data augmentation techniques are used.  ...  Medical image classification and segmentation tasks using deep learning often have limited data due to scarcity of annotated data.  ... 
arXiv:1805.11504v1 fatcat:cv2vkmivb5hedlf35g2qamwugq

A Review of Generative Adversarial Networks in Cancer Imaging: New Applications, New Solutions [article]

Richard Osuala, Kaisar Kushibar, Lidia Garrucho, Akis Linardos, Zuzanna Szafranowska, Stefan Klein, Ben Glocker, Oliver Diaz, Karim Lekadir
2021 arXiv   pre-print
The recent advancements in Generative Adversarial Networks (GANs) in computer vision as well as in medical imaging may provide a basis for enhanced capabilities in cancer detection and analysis.  ...  In this review, we assess the potential of GANs to address a number of key challenges of cancer imaging, including data scarcity and imbalance, domain and dataset shifts, data access and privacy, data  ...  CycleGAN based data augmentation has been shown to be useful for segmentation model training, in particular, for generating images with different acquisition characteristics such as contrast enhanced MRI  ... 
arXiv:2107.09543v1 fatcat:jz76zqklpvh67gmwnsdqzgq5he

Learning More with Less: GAN-based Medical Image Augmentation [article]

Changhee Han, Kohei Murao, Shin'ichi Satoh, Hideki Nakayama
2019 arXiv   pre-print
In Medical Imaging, however, both obtaining medical data and annotating them by expert physicians are challenging; to overcome this lack of data, Data Augmentation (DA) using Generative Adversarial Networks  ...  As a tutorial, this paper introduces GAN-based Medical Image Augmentation, along with tricks to boost classification/object detection/segmentation performance using them, based on our experience and related  ...  Res. 9: 2579-2605, 2008 [13] Han C, Murao K, Noguchi T et al: Learning more with less: conditional PGGAN-based data augmentation for brain metastases detection using highly-rough annotation on MR images  ... 
arXiv:1904.00838v3 fatcat:imzxlzbkirdkhmenhf2kxnb7ly

Robustifying deep networks for image segmentation [article]

Zheng Liu, Jinnian Zhang, Varun Jog, Po-Ling Loh, Alan B McMillan
2019 arXiv   pre-print
Additionally, we explored the effectiveness of distillation and adversarial training via data augmentation to counteract adversarial attacks.  ...  With an increasing interest in applying deep learning techniques to medical imaging data, it is important to quantify the ramifications of adversarial inputs (either intentional or unintentional).  ...  Acknowledgements Research reported in this publication was supported by the National Institute of Biomedical Imaging and Bioengineering of the National Institutes of Health under award number R01EB026708  ... 
arXiv:1908.00656v1 fatcat:7rrmqdst7nfdxniennru66c374
« Previous Showing results 1 — 15 out of 1,285 results