Progres of deep learning based quantitative MRI reconstruction methods
- Vol. 28, Issue 6, Pages: 1698-1708(2023)
Received:18 January 2023,
Revised:28 February 2023,
Published:16 June 2023
DOI: 10.11834/jig.230041
移动端阅览

浏览全部资源
扫码关注微信
Received:18 January 2023,
Revised:28 February 2023,
Published:16 June 2023
移动端阅览
磁共振成像是一种应用广泛的无创医学成像方法,因其丰富的软组织对比度可以成像人体几乎所有内部结构,包括器官、骨骼、肌肉和血管,已成为临床医学影像诊断的利器。然而磁共振成像存在两大公认的瓶颈:成像速度慢、扫描操作烦琐。深度学习给磁共振成像带来莫大的契机,给传统磁共振加速成像带来新的可能。鉴于该领域的快速发展性质,本文旨在总结文献中报道的大量深度学习和磁共振图像重建相结合的方法,以更好地了解该领域。本文简单介绍磁共振成像在多通道线圈接收的并行加速和压缩感知加速下的深度学习重建方法,其中单对比度图像可通过多通道接收线圈提供的冗余度用于加速,多对比度图像可额外使用不同对比度图像这一维度用于加速,动态图像与多对比度图像类似可额外使用时间维度用于加速,本文也将介绍深度学习在这些方面的发展。随着磁共振成像近年来由定性多对比度成像向定量多参数成像的发展,其中定量成像的图像中可内含多对比度图像,如何借用深度学习提供的能力将定性多对比度图像映射到参数图也是一个难点,近年来这一方向也获得了较快的发展,文中也将针对这方面内容进行调研并介绍。针对上述内容,分别介绍国际研究现状和国内研究现状,拟更好地总结国内外研究的发展的异同和趋势。最后对深度学习助力定量磁共振成像方面进行了展望。
Magnetic resonance imaging (MRI) is a commonly-used non-invasive medical imaging method. Due to richable soft tissue-related contrast features of human body, medical imaging diagnosis is beneficial from more internal structures-contextual image in clinic, including its organs, bones, muscles, and blood vessels. However, two sorts of bottlenecks of MRI like slow scanning speed and labor-intensive scanning operation are required to be resolved beyond the constraints of hardware and existing techniques. Current MRI has been facilitating in terms of the emerging deep learning based medical imaging technique. To optimize its acquisition time, parallel imaging and compressed sensing combined with the use of multi-array coils, conventional MRI mainly depends on the hardware improvements and new compactable sequence design. First, literature review of deep learning-based MR reconstruction methods are analyzed, including such deep learning issues originated from 1) integration of parallel imaging and compressed sensing of MRI and 2) acceleration of multi-contrast images and dynamic images of single-contrast image. Deep learning-based reconstruction models is oriented to get their generalizability incorporated with multiple datasets, in which MRI data are challenged for multiple factors like its relevance of centers, models, and field strengths. Two decadal development of quantitative MRI technique can provide pixel-level characterization of intrinsic tissue quantitative parameters, such as T1, T2, and ADC. Generally, quantitative MRI is often required to get multiple weighted images under different parameters, and quantitative tissue parameters can be generated via signal model-required pixel-level nonlinear fitting. Compared to the acquisition of contrast images, acquisition and reconstruction duration has become more longer. To accelerate the acquisition and reconstruction or facilitate accurate mapping, extension ability of deep learning has been developing for the reconstruction of MRI fast imaging. Apart from conventional single parameter mapping method, simultaneous multi-parameter mapping techniques have leaked out, such as MR fingerprinting. Compared to single parameter mapping, multi-parameter mapping can be synchronized and accelerated, and more efficient acquisition and inherent coregistration can be used to deal with such complicated reconstruction. Deep learning technique can be as a multifaceted tool to simplify the reconstruction and speed up the acquisition. However, it is still challenged for such constraints like deep learning algorithms and large amounts of training data. Deep learning methods are required to be melted into magnetic resonance physical models and traditional reconstruction algorithms to a certain extent, and to improve the interpretability of the model and alleviate the impact on large datasets, such methods of data enhancement, weakly supervised learning, unsupervised learning, and transfer learning can be involved in as well. A quantitative MRI-relevant unsupervised network can be potentially used for training in terms of different parameter mapping sequences, which can greatly shrink the amount of data samples further.
Akçakaya M , Moeller S , Weingärtner S and Uğurbil K . 2019 . Scan-specific robust artificial-neural-networks for k-space interpolation (RAKI) reconstruction: database-free deep learning for fast imaging . Magnetic Resonance in Medicine , 81 ( 1 ): 439 - 453 [ DOI: 10.1002/mrm.27420 http://dx.doi.org/10.1002/mrm.27420 ]
Cai C B , Wang C , Zeng Y Q , Cai S H , Liang D , Wu Y W , Chen Z , Ding X H and Zhong J H . 2018 . Single-shot T 2 mapping using overlapping-echo detachment planar imaging and a deep convolutional neural network . Magnetic Resonance in Medicine , 80 ( 5 ): 2202 - 2214 [ DOI: 10.1002/mrm.27205 http://dx.doi.org/10.1002/mrm.27205 ]
Cai C B , Zeng Y Q , Zhuang Y C , Cai S H , Chen L , Ding X H , Bao L J , Zhong J H and Chen Z . 2017 . Single-shot T 2 mapping through overlapping-echo detachment (OLED) planar imaging . IEEE Transactions on Biomedical Engineering , 64 ( 10 ): 2450 - 2461 [ DOI: 10.1109/TBME.2017.2661840 http://dx.doi.org/10.1109/TBME.2017.2661840 ]
Cao X Z , Wang K , Liao C Y , Zhang Z J , Iyer S S , Chen Z F , Lo W C , Liu H F , He H J , Setsompop K , Zhong J H and Bilgic B . 2021 . Efficient T 2 mapping with blip-up/down EPI and gSlider-SMS (T 2 -BUDA-gSlider) . Magnetic Resonance in Medicine , 86 ( 4 ): 2064 - 2075 [ DOI: 10.1002/mrm.28872 http://dx.doi.org/10.1002/mrm.28872 ]
Chen Y , Fang Z H , Hung S C , Chang W T , Shen D G and Lin W L . 2020 . High-resolution 3D MR fingerprinting using parallel imaging and deep learning . NeuroImage , 206 : # 116329 [ DOI: 10.1016/j.neuroimage.2019.116329 http://dx.doi.org/10.1016/j.neuroimage.2019.116329 ]
Chen Y H , Shaw J L , Xie Y B , Li D B and Christodoulou A G . 2019 . Deep learning within a priori temporal feature spaces for large-scale dynamic MR image reconstruction: application to 5-D cardiac MR Multitasking // Proceedings of the 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention . Shenzhen, China : Springer: 495 - 504 [ DOI: 10.1007/978-3-030-32245-8_55 http://dx.doi.org/10.1007/978-3-030-32245-8_55 ]
Chen Y T , Schönlieb C B , Liò P , Leiner T , Dragotti P L , Wang G , Rueckert D , Firmin D and Yang G . 2022 . AI-based reconstruction for fast MRI — a systematic review and meta-analysis . Proceedings of the IEEE , 110 ( 2 ): 224 - 245 [ DOI: 10.1109/JPROC.2022.3141367 http://dx.doi.org/10.1109/JPROC.2022.3141367 ]
Christodoulou A G , Shaw J L , Nguyen C , Yang Q , Xie Y B , Wang N and Li D B . 2018 . Magnetic resonance multitasking for motion-resolved quantitative cardiovascular imaging . Nature Biomedical Engineering , 2 ( 4 ): 215 - 226 [ DOI: 10.1038/s41551-018-0217-y http://dx.doi.org/10.1038/s41551-018-0217-y ]
Du T M , Zhang H G , Li Y M , Pickup S , Rosen M , Zhou R , Song H K and Fan Y . 2021 . Adaptive convolutional neural networks for accelerating magnetic resonance imaging via k-space data interpolation . Medical Image Analysis , 72 : # 102098 [ DOI: 10.1016/j.media.2021.102098 http://dx.doi.org/10.1016/j.media.2021.102098 ]
Eo T , Jun Y , Kim T , Jang J , Lee H J and Hwang D . 2018 . KIKI-net: cross-domain convolutional neural networks for reconstructing undersampled magnetic resonance images . Magnetic Resonance in Medicine , 80 ( 5 ): 2188 - 2201 [ DOI: 10.1002/mrm.27201 http://dx.doi.org/10.1002/mrm.27201 ]
Eo T , Shin H , Jun Y , Kim T and Hwang D . 2020 . Accelerating Cartesian MRI by domain-transform manifold learning in phase-encoding direction . Medical Image Analysis , 63 : # 101689 [ DOI: 10.1016/j.media.2020.101689 http://dx.doi.org/10.1016/j.media.2020.101689 ]
Fang Z H , Chen Y , Liu M X , Xiang L , Zhang Q , Wang Q , Lin W L and Shen D G . 2019 . Deep learning for fast and spatially constrained tissue quantification from highly accelerated data in magnetic resonance fingerprinting . IEEE Transactions on Medical Imaging , 38 ( 10 ): 2364 - 2374 [ DOI: 10.1109/TMI.2019.2899328 http://dx.doi.org/10.1109/TMI.2019.2899328 ]
Feng L , Ma D and Liu F . 2022 . Rapid MR relaxometry using deep learning: an overview of current techniques and emerging trends . NMR in Biomedicine , 35 ( 4 ): # 4416 [ DOI: 10.1002/nbm.4416 http://dx.doi.org/10.1002/nbm.4416 ]
Han Y , Sunwoo L and Ye J C . 2020 . k -space deep learning for accelerated MRI . IEEE Transactions on Medical Imaging , 39 ( 2 ): 377 - 386 [ DOI: 10.1109/TMI.2019.2927101 http://dx.doi.org/10.1109/TMI.2019.2927101 ]
Hoppe E , Körzdörfer G , Würfl T , Wetzl J , Lugauer F , Pfeuffer J and Maier A . 2017 . Deep learning for magnetic resonance fingerprinting: a new approach for predicting quantitative parameter values from time series . Studies in Health Technology and Informatics , 243 : 202 - 206 [ DOI: 10.3233/978-1-61499-808-2-202 http://dx.doi.org/10.3233/978-1-61499-808-2-202 ]
Kang B , Kim B , Schär M , Park H and Heo H Y . 2021 . Unsupervised learning for magnetization transfer contrast MR fingerprinting: application to CEST and nuclear Overhauser enhancement imaging . Magnetic Resonance in Medicine , 85 ( 4 ): 2040 - 2054 [ DOI: 10.1002/mrm.28573 http://dx.doi.org/10.1002/mrm.28573 ]
Khajehim M , Christen T , Tam F and Graham S J . 2021 . Streamlined magnetic resonance fingerprinting: fast whole-brain coverage with deep-learning based parameter estimation . NeuroImage , 238 : # 118237 [ DOI: 10.1016/j.neuroimage.2021.118237 http://dx.doi.org/10.1016/j.neuroimage.2021.118237 ]
Li S M , Wu J , Ma L C , Cai S H and Cai C B . 2022 . A simultaneous multi-slice T 2 mapping framework based on overlapping-echo detachment planar imaging and deep learning reconstruction . Magnetic Resonance in Medicine , 87 ( 5 ): 2239 - 2253 [ DOI: 10.1002/MRM.29128 http://dx.doi.org/10.1002/MRM.29128 ]
Liang D , Cheng J , Ke Z W and Ying L . 2020 . Deep magnetic resonance image reconstruction: inverse problems meet neural networks . IEEE Signal Processing Magazine , 37 ( 1 ): 141 - 151 [ DOI: 10.1109/MSP.2019.2950557 http://dx.doi.org/10.1109/MSP.2019.2950557 ]
Liao C Y , Bilgic B , Tian Q Y , Stockmann J P , Cao X Z , Fan Q Y , Iyer S S , Wang F Y X , Ngamsombat C , Lo W C , Manhard M K , Huang S Y , Wald L L and Setsompop K . 2021 . Distortion-free, high-isotropic-resolution diffusion MRI with gSlider BUDA-EPI and multicoil dynamic B 0 shimming . Magnetic Resonance in Medicine , 86 ( 2 ): 791 - 803 [ DOI: 10.1002/mrm.28748 http://dx.doi.org/10.1002/mrm.28748 ]
Liu F , Feng L and Kijowski R . 2019 . MANTIS: model-augmented neural network with incoherent k-space sampling for efficient MR parameter mapping . Magnetic Resonance in Medicine , 82 ( 1 ): 174 - 188 [ DOI: 10.1002/mrm.27707 http://dx.doi.org/10.1002/mrm.27707 ]
Liu F , Kijowski R , Feng L and El Fakhri G . 2020 . High-performance rapid MR parameter mapping using model-based deep adversarial learning . Magnetic Resonance Imaging , 74 : 152 - 160 [ DOI: 10.1016/j.mri.2020.09.021 http://dx.doi.org/10.1016/j.mri.2020.09.021 ]
Liu F , Kijowski R , El Fakhri G and Feng L . 2021 . Magnetic resonance parameter mapping using model-guided self-supervised deep learning . Magnetic Resonance in Medicine , 85 ( 6 ): 3211 - 3226 [ DOI: 10.1002/mrm.28659 http://dx.doi.org/10.1002/mrm.28659 ]
Ma D , Gulani V , Seiberlich N , Liu K C , Sunshine J L , Duerk J L and Griswold M A . 2013 . Magnetic resonance fingerprinting . Nature , 495 ( 7440 ): 187 - 192 [ DOI: 10.1038/nature11971 http://dx.doi.org/10.1038/nature11971 ]
Ma S , Nguyen C T , Han F , Wang N , Deng Z X , Binesh N , Moser F G , Christodoulou A G and Li D B . 2020 . Three-dimensional simultaneous brain T 1 , T 2 , and ADC mapping with MR multitasking . Magnetic Resonance in Medicine , 84 ( 1 ): 72 - 88 [ DOI: 10.1002/mrm.28092 http://dx.doi.org/10.1002/mrm.28092 ]
Ma S , Wang N , Xie Y B , Fan Z Y , Li D B and Christodoulou A G . 2022 . Motion-robust quantitative multiparametric brain MRI with motion-resolved MR multitasking . Magnetic Resonance in Medicine , 87 ( 1 ): 102 - 119 [ DOI: 10.1002/MRM.28959 http://dx.doi.org/10.1002/MRM.28959 ]
Ouyang B Y , Yang Q Z , Wang X Y , He H J , Ma L C , Yang Q Q , Zhou Z H , Cai S H , Chen Z , Wu Z G , Zhong J H and Cai C B . 2022 . Single-shot T 2 mapping via multi-echo-train multiple overlapping-echo detachment planar imaging and multitask deep learning . Medical Physics , 49 ( 11 ): 7095 - 7107 [ DOI: 10.1002/mp.15820 http://dx.doi.org/10.1002/mp.15820 ]
Pal A and Rathi Y . 2022 . A review and experimental evaluation of deep learning methods for MRI reconstruction . The Journal of Machine Learning for Biomedical Imaging , 1 : #001
Qiu S H , Chen Y H , Ma S , Fan Z Y , Moser F G , Maya M M , Christodoulou A G , Xie Y B and Li D B . 2022 . Multiparametric mapping in the brain from conventional contrast-weighted images using deep learning . Magnetic Resonance in Medicine , 87 ( 1 ): 488 - 495 [ DOI: 10.1002/mrm.28962 http://dx.doi.org/10.1002/mrm.28962 ]
Qu W Y , Cheng J , Zhu Y J and Liang D . 2023 . Deep MR parametric imaging with the learned L + S model and attention mechanism . IET Image Processing , 17 ( 4 ): 969 - 978 [ DOI: 10.1049/ipr2.12687 http://dx.doi.org/10.1049/ipr2.12687 ]
So S , Park H W , Kim B , Fritz F J , Poser B A , Roebroeck A and Bilgic B . 2022 . BUDA-MESMERISE: rapid acquisition and unsupervised parameter estimation for T 1 , T 2 , M 0 , B 0 , and B 1 maps. Magnetic Resonance in Medicine , 88 ( 1 ): 292 - 308 [ DOI: 10.1002/mrm.29228 http://dx.doi.org/10.1002/mrm.29228 ]
Sun L Y , Fan Z W , Fu X Y , Huang Y , Ding X H and Paisley J . 2019 . A deep information sharing network for multi-contrast compressed sensing MRI reconstruction . IEEE Transactions on Image Processing , 28 ( 12 ): 6141 - 6153 [ DOI: 10.1109/TIP.2019.2925288 http://dx.doi.org/10.1109/TIP.2019.2925288 ]
Wang C , Wu Y W , Ding X H , Huang Y and Cai C B . 2018 . High efficient reconstruction of single-shot magnetic resonance T 2 mapping through overlapping echo detachment and DenseNet // Proceedings of the 25th International Conference on Neural Information Processing . Siem Reap, Cambodia : Springer: 408 - 418 [ DOI: 10.1007/978-3-030-04224-0_35 http://dx.doi.org/10.1007/978-3-030-04224-0_35 ]
Wang H F , Cheng J , Jia S , Qiu Z L , Shi C Y , Zou L X , Su S , Chang Y C , Zhu Y J , Ying L and Liang D . 2019 . Accelerating MR imaging via deep chambolle-pock network // Proceedings of the 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society . Berlin, Germany : IEEE: 6818 - 6821 [ DOI: 10.1109/EMBC.2019.8857141 http://dx.doi.org/10.1109/EMBC.2019.8857141 ]
Wang S S , Cheng H T , Ying L , Xiao T H , Ke Z W , Zheng H R and Liang D . 2020 . DeepcomplexMRI: exploiting deep residual network for fast parallel MR imaging with complex convolution . Magnetic Resonance Imaging , 68 : 136 - 147 [ DOI: 10.1016/j.mri.2020.02.002 http://dx.doi.org/10.1016/j.mri.2020.02.002 ]
Wang S S , Ke Z W , Cheng H T , Jia S , Ying L , Zheng H R and Liang D . 2022 . DIMENSION: dynamic MR imaging with both k-space and spatial prior knowledge obtained via multi-supervised network training . NMR in Biomedicine , 35 ( 4 ): #e 4131 [ DOI: 10.1002/nbm.4131 http://dx.doi.org/10.1002/nbm.4131 ]
Wang S S , Su Z H , Ying L , Peng X , Zhu S , Liang F , Feng D G and Liang D . 2016 . Accelerating magnetic resonance imaging via deep learning // The 13th International Symposium on Biomedical Imaging . Prague, Czech Republic : IEEE: 514 - 517 [ DOI: 10.1109/ISBI.2016.7493320 http://dx.doi.org/10.1109/ISBI.2016.7493320 ]
Wang S S , Xiao T H , Liu Q G and Zheng H R . 2021 . Deep learning for fast MR imaging: a review for learning reconstruction from incomplete k-space data . Biomedical Signal Processing and Control , 68 : # 102579 [ DOI: 10.1016/j.bspc.2021.102579 http://dx.doi.org/10.1016/j.bspc.2021.102579 ]
Wang Z , Qian C , Guo D , Sun H W , Li R S , Zhao B and Qu X B . 2023 . One-dimensional deep low-rank and sparse network for accelerated MRI . IEEE Transactions on Medical Imaging , 42 ( 1 ): 79 - 90 [ DOI: 10.1109/TMI.2022.3203312 http://dx.doi.org/10.1109/TMI.2022.3203312 ]
Xiang L , Chen Y , Chang W T , Zhan Y Q , Lin W L , Wang Q and Shen D G . 2019 . Deep-learning-based multi-modal fusion for fast MR reconstruction . IEEE Transactions on Biomedical Engineering , 66 ( 7 ): 2105 - 2114 [ DOI: 10.1109/TBME.2018.2883958 http://dx.doi.org/10.1109/TBME.2018.2883958 ]
Yang M R , Jiang Y , Ma D , Mehta B B and Griswold M A . 2020 . Game of learning bloch equation simulations for MR fingerprinting [EB/OL]. [ 2023-01-18 ]. https://arxiv.org/pdf/2004.02270.pdf https://arxiv.org/pdf/2004.02270.pdf
Zeng G S , Guo Y , Zhan J Y , Wang Z , Lai Z Y , Du X F , Qu X B and Guo D . 2021 . A review on deep learning MRI reconstruction without fully sampled k-space . BMC Medical Imaging , 21 ( 1 ): # 195 [ DOI: 10.1186/S12880-021-00727-9 http://dx.doi.org/10.1186/S12880-021-00727-9 ]
Zeng W , Peng J , Wang S S and Liu Q G . 2020 . A comparative study of CNN-based super-resolution methods in MRI reconstruction and its beyond . Signal Processing: Image Communication , 81 : # 115701 [ DOI: 10.1016/j.image.2019.115701 http://dx.doi.org/10.1016/j.image.2019.115701 ]
Zhang J , Wu J , Chen S J , Zhang Z Y , Cai S H , Cai C B and Chen Z . 2019 . Robust single-shot T 2 mapping via multiple overlapping-echo acquisition and deep neural network . IEEE Transactions on Medical Imaging , 38 ( 8 ): 1801 - 1811 [ DOI: 10.1109/TMI.2019.2896085 http://dx.doi.org/10.1109/TMI.2019.2896085 ]
Zhang Z J , Cho J , Wang L , Liao C Y , Shin H G , Cao X Z , Lee J , Xu J M , Zhang T , Ye H H , Setsompop K , Liu H F and Bilgic B . 2022 . Blip up-down acquisition for spin-and gradient-echo imaging (BUDA-SAGE) with self-supervised denoising enables efficient T 2 , T 2 *, para-and dia-magnetic susceptibility mapping. Magnetic Resonance in Medicine , 88 ( 2 ): 633 - 650 [ DOI: 10.1002/mrm.29219 http://dx.doi.org/10.1002/mrm.29219 ]
Zhu B , Liu J Z , Cauley S F , Rosen B R and Rosen M S . 2018 . Image reconstruction by domain-transform manifold learning . Nature , 555 ( 7697 ): 487 - 492 [ DOI: 10.1038/nature25988 http://dx.doi.org/10.1038/nature25988 ]
相关文章
相关作者
相关机构
京公网安备11010802024621