通过深度学习加速磁共振成像
ACCELERATING MAGNETIC RESONANCE IMAGING VIA DEEP LEARNING.
作者信息
Wang Shanshan, Su Zhenghang, Ying Leslie, Peng Xi, Zhu Shun, Liang Feng, Feng Dagan, Liang Dong
机构信息
Paul C. Lauterbur Research Center for Biomedical Imaging, SIAT, CAS, Shenzhen, P.R.China.
School of Information Technologies, Guangdong University of Technology, Guangzhou, P.R. China.
出版信息
Proc IEEE Int Symp Biomed Imaging. 2016 Apr;2016:514-517. doi: 10.1109/ISBI.2016.7493320. Epub 2016 Jun 16.
Abstract
This paper proposes a deep learning approach for accelerating magnetic resonance imaging (MRI) using a large number of existing high quality MR images as the training datasets. An off-line convolutional neural network is designed and trained to identify the mapping relationship between the MR images obtained from zero-filled and fully-sampled k-space data. The network is not only capable of restoring fine structures and details but is also compatible with online constrained reconstruction methods. Experimental results on real MR data have shown encouraging performance of the proposed method for efficient and effective imaging.
摘要
本文提出了一种深度学习方法，该方法利用大量现有的高质量磁共振成像（MRI）图像作为训练数据集来加速MRI。设计并训练了一个离线卷积神经网络，以识别从零填充和全采样k空间数据获得的MR图像之间的映射关系。该网络不仅能够恢复精细结构和细节，还与在线约束重建方法兼容。对真实MR数据的实验结果表明，所提出的方法在高效成像方面具有令人鼓舞的性能。
相似文献
1
ACCELERATING MAGNETIC RESONANCE IMAGING VIA DEEP LEARNING.通过深度学习加速磁共振成像
Proc IEEE Int Symp Biomed Imaging. 2016 Apr;2016:514-517. doi: 10.1109/ISBI.2016.7493320. Epub 2016 Jun 16.
2
Improving resolution of MR images with an adversarial network incorporating images with different contrast.利用具有不同对比度的图像的对抗网络提高磁共振图像的分辨率。
Med Phys. 2018 Jul;45(7):3120-3131. doi: 10.1002/mp.12945. Epub 2018 May 18.
3
MRI Gibbs-ringing artifact reduction by means of machine learning using convolutional neural networks.基于卷积神经网络的机器学习方法降低 MRI 的 Gibbs 环伪影。
Magn Reson Med. 2019 Dec;82(6):2133-2145. doi: 10.1002/mrm.27894. Epub 2019 Aug 2.
4
MR-based synthetic CT generation using a deep convolutional neural network method.基于磁共振成像利用深度卷积神经网络方法生成合成CT图像
Med Phys. 2017 Apr;44(4):1408-1419. doi: 10.1002/mp.12155. Epub 2017 Mar 21.
5
SelfCoLearn: Self-Supervised Collaborative Learning for Accelerating Dynamic MR Imaging.SelfCoLearn：用于加速动态磁共振成像的自监督协作学习
Bioengineering (Basel). 2022 Nov 4;9(11):650. doi: 10.3390/bioengineering9110650.
6
Shape constrained fully convolutional DenseNet with adversarial training for multiorgan segmentation on head and neck CT and low-field MR images.基于对抗训练的形状约束全卷积 DenseNet 用于头颈部 CT 和低场 MR 图像多器官分割。
Med Phys. 2019 Jun;46(6):2669-2682. doi: 10.1002/mp.13553. Epub 2019 May 6.
7
Region-of-interest undersampled MRI reconstruction: A deep convolutional neural network approach.感兴趣区域欠采样 MRI 重建：一种深度卷积神经网络方法。
Magn Reson Imaging. 2019 Nov;63:185-192. doi: 10.1016/j.mri.2019.07.010. Epub 2019 Jul 25.
8
MANTIS: Model-Augmented Neural neTwork with Incoherent k-space Sampling for efficient MR parameter mapping.MANTIS：用于高效磁共振参数映射的基于模型增强的神经网络与非相干 k 空间采样。
Magn Reson Med. 2019 Jul;82(1):174-188. doi: 10.1002/mrm.27707. Epub 2019 Mar 12.
9
Motion Artifact Reduction Using a Convolutional Neural Network for Dynamic Contrast Enhanced MR Imaging of the Liver.基于卷积神经网络的动态对比增强磁共振成像肝脏运动伪影减少。
Magn Reson Med Sci. 2020 Feb 10;19(1):64-76. doi: 10.2463/mrms.mp.2018-0156. Epub 2019 Apr 26.
10
Fast learning of fiber orientation distribution function for MR tractography using convolutional neural network.利用卷积神经网络快速学习磁共振纤维追踪的纤维方向分布函数。
Med Phys. 2019 Jul;46(7):3101-3116. doi: 10.1002/mp.13555. Epub 2019 May 11.
引用本文的文献
1
CoRRECT: A Deep Unfolding Framework for Motion-Corrected Quantitative R2* Mapping.正确：用于运动校正定量R2*映射的深度展开框架。
J Math Imaging Vis. 2025 Apr;67(2). doi: 10.1007/s10851-025-01236-y. Epub 2025 Apr 2.
2
Enhanced Partial Fourier MRI With Zero-Shot Deep Untrained Priors.基于零样本深度未训练先验的增强部分傅里叶磁共振成像
IEEE Access. 2024;12:182157-182170. doi: 10.1109/access.2024.3508761. Epub 2024 Nov 29.
3
Diff5T: Benchmarking human brain diffusion MRI with an extensive 5.0 Tesla k-space and spatial dataset.Diff5T：使用广泛的5.0特斯拉k空间和空间数据集对人类脑扩散磁共振成像进行基准测试。
Sci Data. 2025 Aug 4;12(1):1352. doi: 10.1038/s41597-025-05640-2.
4
Unsupervised single-image super-resolution for infant brain MRI.用于婴儿脑部磁共振成像的无监督单图像超分辨率技术
Neuroimage. 2025 Jun 21;317:121293. doi: 10.1016/j.neuroimage.2025.121293.
5
Learning Task-Specific Strategies for Accelerated MRI.学习用于加速磁共振成像的特定任务策略。
IEEE Trans Comput Imaging. 2024;10:1040-1054. doi: 10.1109/tci.2024.3410521. Epub 2024 Jul 1.
6
Diffusion probabilistic generative models for accelerated, in-NICU permanent magnet neonatal MRI.用于加速的新生儿重症监护病房永磁新生儿MRI的扩散概率生成模型。
Magn Reson Med. 2025 Oct;94(4):1546-1562. doi: 10.1002/mrm.30585. Epub 2025 Jun 17.
7
Robust multi-coil MRI reconstruction via self-supervised denoising.通过自监督去噪实现稳健的多线圈磁共振成像重建
Magn Reson Med. 2025 Jun 2. doi: 10.1002/mrm.30591.
8
Deep Learning-Based Reconstruction for Accelerated Cervical Spine MRI: Utility in the Evaluation of Myelopathy and Degenerative Diseases.基于深度学习的加速颈椎磁共振成像重建：在脊髓病和退行性疾病评估中的应用
AJNR Am J Neuroradiol. 2025 Apr 2;46(4):750-757. doi: 10.3174/ajnr.A8567.
9
Accelerated Abdominal MRI: A Review of Current Methods and Applications.腹部快速磁共振成像：当前方法与应用综述
J Magn Reson Imaging. 2025 Sep;62(3):654-672. doi: 10.1002/jmri.29750. Epub 2025 Mar 18.
10
Reconstruction techniques for accelerating dynamic cardiovascular magnetic resonance imaging.加速动态心血管磁共振成像的重建技术
J Cardiovasc Magn Reson. 2025 Mar 6;27(1):101873. doi: 10.1016/j.jocmr.2025.101873.
本文引用的文献
1
Fenchel duality based dictionary learning for restoration of noisy images.基于 Fenchel 对偶字典学习的噪声图像恢复。
IEEE Trans Image Process. 2013 Dec;22(12):5214-25. doi: 10.1109/TIP.2013.2282900.
2
Highly undersampled magnetic resonance image reconstruction using two-level Bregman method with dictionary updating.基于字典更新的两级 Bregman 方法的高欠采样磁共振图像重建。
IEEE Trans Med Imaging. 2013 Jul;32(7):1290-301. doi: 10.1109/TMI.2013.2256464. Epub 2013 Apr 2.
3
3D convolutional neural networks for human action recognition.三维卷积神经网络的人体动作识别。
IEEE Trans Pattern Anal Mach Intell. 2013 Jan;35(1):221-31. doi: 10.1109/TPAMI.2012.59.
4
Manifold learning for image-based breathing gating in ultrasound and MRI.基于图像的超声和 MRI 呼吸门控的流形学习。
Med Image Anal. 2012 May;16(4):806-18. doi: 10.1016/j.media.2011.11.008. Epub 2011 Dec 8.
5
Parallel imaging with nonlinear reconstruction using variational penalties.基于变分罚函数的非线性重建并行成像。
Magn Reson Med. 2012 Jan;67(1):34-41. doi: 10.1002/mrm.22964. Epub 2011 Jun 27.
6
Multivariate compressive sensing for image reconstruction in the wavelet domain: using scale mixture models.基于尺度混合模型的小波域图像重建的多元压缩感知。
IEEE Trans Image Process. 2011 Dec;20(12):3483-94. doi: 10.1109/TIP.2011.2150231. Epub 2011 May 5.
7
Fast MR image reconstruction for partially parallel imaging with arbitrary k-space trajectories.任意 k 空间轨迹的部分并行成像快速磁共振图像重建。
IEEE Trans Med Imaging. 2011 Mar;30(3):575-85. doi: 10.1109/TMI.2010.2088133.
8
Accelerated dynamic MRI exploiting sparsity and low-rank structure: k-t SLR.利用稀疏性和低秩结构的加速动态 MRI：k-t SLR。
IEEE Trans Med Imaging. 2011 May;30(5):1042-54. doi: 10.1109/TMI.2010.2100850. Epub 2011 Jan 31.
9
Computational acceleration for MR image reconstruction in partially parallel imaging.部分并行成像中磁共振图像重建的计算加速。
IEEE Trans Med Imaging. 2011 May;30(5):1055-63. doi: 10.1109/TMI.2010.2073717. Epub 2010 Sep 7.
10
SPIRiT: Iterative self-consistent parallel imaging reconstruction from arbitrary k-space.SPIRiT：任意 k 空间的迭代自一致并行成像重建。
Magn Reson Med. 2010 Aug;64(2):457-71. doi: 10.1002/mrm.22428.
Zotero 插件