分享:
分享到微信朋友圈
X
综述
深度学习在化学交换饱和转移磁共振成像中的研究进展
张利红 许崇欣 侯蓓蓓 唐朝生 孙君顶

Cite this article as: Zhang LH, Xu CX, Hou BB, et al. Research progress of deep learning in chemical exchange saturation transfer magnetic resonance imaging[J]. Chin J Magn Reson Imaging, 2022, 13(11): 165-168.本文引用格式:张利红, 许崇欣, 侯蓓蓓, 等. 深度学习在化学交换饱和转移磁共振成像中的研究进展[J]. 磁共振成像, 2022, 13(11): 165-168. DOI:10.12015/issn.1674-8034.2022.11.034.


[摘要] 化学交换饱和转移(chemical exchange saturation transfer, CEST)是一种新型的MRI技术,其基本原理是通过水信号的减少来间接实现对特定低浓度溶质分子的检测。采集速度慢、量化速度慢、量化评估不准确等问题影响着CEST MRI在临床中的应用推广,如何改善这些问题也成为研究的重点。深度学习作为人工智能的一种新的研究方向,近几年才应用于CEST MRI技术。本文在广泛调研国内外文献的基础上,对深度学习在临床CEST MRI上应用进行了深入分析与梳理。其中,在量化方面,一方面介绍了通过给深度神经网络(deep neural network, DNN)中输入临床中采集3 T的Z谱数据,预测出高场的CEST参数,进而得到比较明显的CEST信号;另一方面介绍了DNN结合磁化转移指纹识别(magnetization transfer fingerprinting, MTF)技术的方法改善传统量化方法中拟合参数精度低和拟合效率低的问题;在加速方面,一方面介绍深度学习用于CEST MRI加速采集;另一方面介绍了深度学习用于改善传统多池洛伦兹拟合量化速度慢的问题。供对本领域感兴趣者参考及在此基础上进一步地研究开发,加速CEST MRI的临床转换。
[Abstract] Deep learning, a significant method of artificial intelligence, has been used for chemical exchange saturation transfer magnetic resonance imaging (CEST MRI) in recent years, the basic principle is to indirectly realize the detection of specific low concentration of solute molecules through the reduction of water signal. Problems such as slow collection speed, slow quantification speed, and inaccurate quantitative evaluation affect the application and promotion of CEST MRI in clinical practice, and how to improve these problems has also become the focus of research. As a new research direction of artificial intelligence, deep learning has only been applied to CEST-MRI technology in recent years. This method is mainly used in the quantification and acceleration aspects of CEST MRI. The quantification usage includes prediction of the high field results and quantify proton exchange rate and concentration. The acceleration studies include acceleration on acquisition and acceleration on quantification. As for the method itself, the most frequently used algorithm is convolutional neural network and deep neural networks. Other studies included the comparison among different deep learning models and establishment of deep learning models based on different MRI sequences. This paper is to review the application of deep learning in CEST MRI in detail,which can be used as reference for interested parties in this field and further research and development on this basis. Then accelerate the clinical transformation of CEST MRI.
[关键词] 磁共振成像;化学交换饱和转移;深度学习;量化;加速
[Keywords] magnetic resonance imaging;chemical exchange saturation transfer;deep learning;quantitation;acceleration

张利红    许崇欣    侯蓓蓓    唐朝生    孙君顶 *  

河南理工大学计算机科学与技术学院,焦作 454003

孙君顶,E-mail:sunjd@hpu.edu.cn

作者利益冲突声明:全体作者均声明无利益冲突。


基金项目: 河南省科技攻关项目 212102310084 河南理工大学博士基金项目 B2022-11
收稿日期:2022-03-31
接受日期:2022-09-14
中图分类号:R445.2 
文献标识码:A
DOI: 10.12015/issn.1674-8034.2022.11.034
本文引用格式:张利红, 许崇欣, 侯蓓蓓, 等. 深度学习在化学交换饱和转移磁共振成像中的研究进展[J]. 磁共振成像, 2022, 13(11): 165-168. DOI:10.12015/issn.1674-8034.2022.11.034.

       深度学习是机器学习研究中的一个新领域[1],是对采用较深的神经网络结构的一大类算法的总称。化学交换饱和转移(chemical exchange saturation transfer, CEST)是磁共振成像领域的一种新型对比增强机制[2, 3, 4],可对生物体内多种代谢物进行活体检测,CEST成像是通过间接测量与饱和溶质质子的化学交换引起的水信号变化来反映特定化学基团的信息[5, 6, 7]。尽管这些小溶质在生物组织中的浓度通常仅在毫摩尔范围内,只要选择合适的实验参数,水与溶质池中饱和质子化学交换的累积效应就能实现信号增强,具有增强的灵敏度。例如蛋白质、脂类、糖原等的CEST MRI技术,灵敏度较磁共振波谱成像(magnetic resonance spectrum, MRS)技术提高了1000倍以上[8]。大量的研究证明该成像技术有潜力应用到临床诊断中,但目前并没有成为临床中常规的影像检查方法[9]。一方面是因为临床CEST MRI采集时间长[10];另一方面是因为目前大多数CEST MRI的成像方案都不能提供组织参数的定量测量[11],仅能得到半固体磁化转移(magnetization transfer, MT)和CEST的加权图像,且各种混杂因素影响其对比加权图像的解释[5,12, 13, 14, 15]。因此如何将先进的人工智能技术更好地应用于CEST MRI的采集和量化中,是当前该研究中的挑战和热点之一[16]。研究表明:深度学习可以应用于CEST MRI的采集加速和谱分析中[16],深度学习应用CEST MRI的临床意义为:(1)在常规相对静止的扫描部位,该研究可缩短扫描时间,减少接受扫描检查者由于长时间扫描导致的不舒适感,提高了检查的成功率;(2)对于MRI扫描运动的部位,该研究可缩短扫描时间,让受检查者以一定的状态配合完成检查,例如心脏或腹部扫描时,通过缩短序列的扫描时间能够让受检查者在一次屏气的状态下完成一个序列的检查,进而拓宽了CEST MRI的应用领域,从相对静止的部位到运动配合的部位;(3)该研究可大大缩短3D检查序列的采集时间,拓宽MRI的临床应用,例如应用CEST MRI进行增强血管成像;(4)该研究能够在相同时间分辨率的前提下提高图像的空间分辨率,使得在控制CEST MRI成像时间的情况下获得高分辨率的图像。

1 深度学习应用于CEST MRI的采集

       CEST MRI需要采集一系列频偏的CEST图像,因此采集时间比较长[17, 18],对于临床3D的序列,这个问题将更加突出[19, 20]。对于活体成像,采集时间长还容易引起运动伪影[21],导致诊断误差。因此,如何优化采集序列、加速CEST MRI的采集[22]一直是CEST MRI应用到临床亟待解决的问题[23],尤其是对于临床的采集[9],不同的深度学习方法都对这一问题进行了研究。

       Li等[24]通过结合广泛激活的深度超分辨率网络中的广泛激活残余学习网络这一模块,在B0场不均匀性校正精度不变的情况下,显著减少了不同频率偏移的数量,进而加速采集,并通过每一层都使用卷积神经网络(convolutional neural network, CNN)的方法有效解决了CEST MRI信噪比值较低的问题;Guo等[25]提出通过周期旋转叠加平行线提高重建CEST MRI的方法加速采集,该方法仅采集每个饱和图像的k空间中心的几条线,深度神经网络用于直接从欠采样的饱和图像中量化CEST效应,从而大幅度缩短采集时间;为了提高采集效率和重建精度,Kang等[26]提出了一个基于学习的采集时间表(LOAS)优化框架,以最少的扫描参数优化射频饱和编码的磁化转移指纹识别(magnetization transfer fingerprinting, MTF)采集;此外,Bie等[27]提出使用CNN的分类模型,仿真数据作为训练,在正常采集和稀疏采样两种采集情况下,所提出的分类模型均可以预测出不同类型的乳腺癌组织和正常的肌肉组织,该训练模型可以通过稀疏采样加速采集。

       以上研究表明:不同的深度神经网络框架和策略尝试加速采集均大幅度地缩短了采集时间,但仍存在一些问题有待解决,如没有考虑B1场不均匀性的问题以及分类的性能好坏过度依赖测试数据。

2 深度学习在CEST MRI量化中的应用

2.1 深度学习预测不同场强下的量化参数

       CEST MRI可以通过水信号的减少间接灵敏地检测出特定分子[28],但是水信号的降低不仅包含有用的CEST信号,而且还包含MT、水的直接饱和(direct water saturation)和核奥氏效应(nuclear Overhauser enhancement, NOE)[3,29, 30, 31, 32]。传统提取CEST信号诸如布洛赫拟合、洛伦兹差、多池洛伦兹拟合等量化方法非常耗时[33, 34, 35, 36, 37]。由于CEST和MT交换参数值对诸如癌症、脑卒中以及多发硬化等疾病的诊断都很重要[9,37, 38, 39, 40, 41, 42, 43, 44, 45],因此如何快速得到定量MT和CEST的参数一直是CEST MRI研究的一个热点和难点,可尝试利用深度学习方法解决该问题。

       Glang等[46]提出引入了一种具有不确定性量化的概率输出层的深度前馈神经网络,快速预测洛伦兹的拟合参数(约1秒)加速临床CEST数据的分析。Huang等[47]又提出通过搭建具有优化的隐藏层和神经元的全连接神经网络模型deepCEST/deepAREX,对野生型小鼠的CEST数据进行反向传播优化训练,可以快速(约1秒)精确地得到临床3 T场强下采集且未经过训练的阿尔茨海默病小鼠的拟合参数;针对人工定位和定量缺血性脑卒中病变区域费时费力、不能满足及时治疗干预的问题,Zhao等[48]提出使用一个基于全卷积神经网络的模型分割脑缺血区域,最终使用Grad-cam方法,快速得到图中每个像素与“缺血”类相关性的粗糙定位图;Guo等[49]提出通过酰胺质子转移加权MRI数据加入到CNN的输入端,可以较为准确地评估出脑胶质瘤的疗效。

       此外,在高场上检测CEST信号比较可靠[35],但是临床3 T采集得到的Z谱数据由于谱峰拓宽,导致提取CEST信号的难度增大[50],这也是CEST MRI研究的一个难点[11]。Zaiss等[50]提出使用3层全连接的神经网络,输入数据为3 T的Z谱数据,预测的目标结果是9.4 T的CEST拟合参数,虽然这种方法的临床应用还需要进一步的验证,但该方法可以帮助医生决定临床中哪些患者需要进一步用超高频的CEST扫描。

       几种不同的深度学习网络框架均可以预测出量化的参数,但是不同饱和功率的测试需要进一步验证,对于临床数据也需要进一步的测试验证。

2.2 深度学习量化质子的交换率、浓度和频偏等参数

       MTF技术是从采集CEST谱中量化出MT和CEST参数的一种常用的量化方法[12,22,26,51,52]。但是,传统的MTF一方面由于通过求解布洛赫方程(Maxwell-Bloch equation)的数值解建立字典的方法比较耗时,另一方面用于定量MT和CEST的图像的重建方法也比较耗时[51],这是MTF需要迫切解决的问题[53],考虑到MTF可以通过人工合成大量数据解决深度学习数据不足的问题,深度学习可尝试解决MTF重建MT和CEST参数的问题。

       Kim等[52]提出深度神经网络(deep neural network, DNN)结合MTF技术的方法,该方法在临床可接受的扫描时间内,定量出用于重建3D的MT、CEST和NOE图像半固体MT的交换率以及T1、T2等参数,但该方法需要比较长的优化和训练时间;此外,针对传统的监督学习标注数据比较耗时和训练结果受到标注准确性影响的问题,Kang等[54]提出无监督的CNN定量方法,通过定义实验数据与合成MTF数据作为损失函数,迭代的优化训练达到实验数据与合成MTF数据之间的最小,最终预测出低浓度下量化CEST的交换率、浓度等参数;Karunanithy等[55]提出利用DNN网络在不需要调整额外参数的前提下可以较为准确地预测出CEST的频偏;Perlman等[10]提出使用一个已经应用到MRI的深度神经网络AutoCEST,该网络由一个CEST饱和模块、一个自旋动力学模块和一个深度重构网络组成,每个模块均可微调且每个模块连接在一起,最终得到量化CEST MRI的浓度和交换率等参数;此外,Chen等[56]结合在隐藏层包含足够多的神经元的全连接前馈神经网络非常适合找到一维数据里的特征这个特点,提出在临床的场强下基于化学交换饱和转移人工神经网络(artificial neural network-based chemical exchange saturation transfer, ANNCEST)的方法,通过使用BM方程生成的Z谱训练搭建的网络,训练提取出Z谱数据和量化参数之间的关系,在训练好的ANNCEST框架下很快(约几秒)定量出人体骨骼肌的磷酸肌酸交换参数,由于该方法使用变化的CEST编码作为输入,和MTF比较类似,因此这种方法被视为CEST-MTF的变体[20];此外,Karunanithy等[55]提出通过DNN网络预测CEST的化学位移。

       深度神经网络作为机器学习的一个新的分支,在量化CEST MRI中质子的交换率、浓度和频偏等参数逐渐展示其价值。但是,仍有部分问题需要进一步的研究,如:质子的量化精度过度依赖于测试数据的问题、参数量化的准确性和鲁棒性问题以及量化精度与复杂度的综合权衡问题等。

3 总结及展望

       目前深度学习在CEST MRI的研究仅是组内的数据,将组内采集的数据分为实验组和测试组,并无公开的外部数据验证,这种深度学习的模式并不完整。在之后的研究中,需要用训练好的模型对其他大量的公开数据进行预测。虽然深度神经网络通常被认为是“黑盒”,即对从输入到预测的映射的洞察力有限,但其在CEST MRI的应用取得了一定的效果。随着CEST MRI和深度学习的进一步发展,深度学习在CEST MRI中应用的广度和深度均得到改善,未来深度学习的方法可能会成为CEST MRI研究的一个有力的工具[11]

[1]
Deng L. Deep Learning: Methods and Applications[J]. Foundations and Trends® in Signal Processing, 2014, 7(3-4): 197-387. DOI: 10.1561/2000000039.
[2]
Zhang LH, Zhao YC, Chen YR, et al. Voxel-wise Optimization of Pseudo Voigt Profile (VOPVP) for Z-spectra fitting in chemical exchange saturation transfer (CEST) MRI[J]. Quant Imaging Med Surg, 2019, 9(10): 1714-1730. DOI: 10.21037/qims.2019.10.01.
[3]
Jiang SS, Wen ZB, Ahn SS, et al. Applications of chemical exchange saturation transfer magnetic resonance imaging in identifying genetic markers in gliomas[J/OL]. NMR Biomed, 2022 [2022-09-12]. https://doi.org/10.1002/nbm.4731. DOI: 10.1002/nbm.4731.
[4]
Ward KM, Aletras AH, Balaban RS. A new class of contrast agents for MRI based on proton chemical exchange dependent saturation transfer (CEST)[J]. J Magn Reson, 2000, 143(1): 79-87. DOI: 10.1006/jmre.1999.1956.
[5]
Zu ZL, Lin EC, Louie EA, et al. Chemical exchange rotation transfer imaging of phosphocreatine in muscle[J/OL]. NMR Biomed, 2021 [2022-09-12]. https://doi.org/10.1002/nbm.4437. DOI: 10.1002/nbm.4437.
[6]
Zhou JY, Lal B, Wilson DA, et al. Amide proton transfer (APT) contrast for imaging of brain tumors[J]. Magn Reson Med, 2003, 50(6): 1120-1126. DOI: 10.1002/mrm.10651.
[7]
Zhou IY, Wang EF, Cheung JS, et al. Quantitative chemical exchange saturation transfer (CEST) MRI of glioma using Image Downsampling Expedited Adaptive Least-squares (IDEAL) fitting[J/OL]. Sci Rep, 2017 [2022-09-12]. https://doi.org/10.1038/s41598-017-00167-y. DOI: 10.1038/s41598-017-00167-y.
[8]
Liu GS, Song XL, Chan KWY, et al. Nuts and bolts of chemical exchange saturation transfer MRI[J]. NMR Biomed, 2013, 26(7): 810-828. DOI: 10.1002/nbm.2899.
[9]
Jones KM, Pollard AC, Pagel MD. Clinical applications of chemical exchange saturation transfer (CEST) MRI[J]. J Magn Reson Imaging, 2018, 47(1): 11-27. DOI: 10.1002/jmri.25838.
[10]
Perlman O, Zhu B, Zaiss M, et al. An end-to-end AI-based framework for automated discovery of rapid CEST/MT MRI acquisition protocols and molecular parameter quantification (AutoCEST)[J/OL]. Magn Reson Med, 2022 [2022-09-12]. https://doi.org/10.1002/mrm.29173. DOI: 10.1002/mrm.29173.
[11]
Huang J, Chen Z, Park SW, et al. Molecular imaging of brain tumors and drug delivery using CEST MRI: promises and challenges[J/OL]. Pharmaceutics, 2022 [2022-09-12]. https://doi.org/10.3390/pharmaceutics14020451. DOI: 10.3390/pharmaceutics14020451.
[12]
Perlman O, Ito H, Herz K, et al. Quantitative imaging of apoptosis following oncolytic virotherapy by magnetic resonance fingerprinting aided by deep learning[J/OL]. Nat Biomed Eng, 2021 [2022-09-12]. https://doi.org/10.1038/s41551-021-00809-7. DOI: 10.1038/s41551-021-00809-7.
[13]
Zu Z. Towards the complex dependence of MTRasym on T1w in amide proton transfer (APT) imaging[J/OL]. NMR Biomed, 2018 [2022-09-12].https://doi.org/10.1002/nbm.3934. DOI: 10.1002/nbm.3934.
[14]
Zu ZL. Toward more reliable measurements of NOE effects in CEST spectra at around-1.6 ppm (NOE (-1.6)) in rat brain[J]. Magn Reson Med, 2019, 81(1): 208-219. DOI: 10.1002/mrm.27370.
[15]
Zhou IY, Lu DS, Ji Y, et al. Determination of multipool contributions to endogenous amide proton transfer effects in global ischemia with high spectral resolution in vivo chemical exchange saturation transfer MRI[J]. Magn Reson Med, 2019, 81(1): 645-652. DOI: 10.1002/mrm.27385.
[16]
van Zijl PCM, Brindle K, Lu HZ, et al. Hyperpolarized MRI, functional MRI, MR spectroscopy and CEST to provide metabolic information in vivo[J]. Curr Opin Chem Biol, 2021, 63: 209-218. DOI: 10.1016/j.cbpa.2021.06.003.
[17]
Zhang Y, Heo HY, Lee DH, et al. Selecting the reference image for registration of CEST series[J]. J Magn Reson Imaging, 2016, 43(3): 756-761. DOI: 10.1002/jmri.25027.
[18]
Zhang Y, Zu T, Liu RB, et al. Acquisition sequences and reconstruction methods for fast chemical exchange saturation transfer imaging[J/OL]. NMR Biomed, 2022 [2022-09-12]. https://doi.org/10.1002/nbm.4699. DOI: 10.1002/nbm.4699.
[19]
Liu R, Zhang H, Niu W, et al. Erratum to: improved chemical exchange saturation transfer imaging with real-time frequency drift correction (Magn Reson Med. 2019; 81: 2915-2923)[J/OL]. Magn Reson Med, 2020 [2022-09-12].https://doi.org/10.1002/mrm.28133. DOI: 10.1002/mrm.28133.
[20]
Wada T, Tokunaga C, Togao O, et al. Three-dimensional chemical exchange saturation transfer imaging using compressed SENSE for full z-spectrum acquisition[J]. Magn Reson Imaging, 2022, 92: 58-66. DOI: 10.1016/j.mri.2022.05.014.
[21]
Bie CX, Liang YH, Zhang LH, et al. Motion correction of chemical exchange saturation transfer MRI series using robust principal component analysis (RPCA) and PCA[J]. Quant Imaging Med Surg, 2019, 9(10): 1697-1713. DOI: 10.21037/qims.2019.09.14.
[22]
Liu J, Liu H, Liu Q, et al. Encoding capability prediction of acquisition schedules in CEST MR fingerprinting for pH quantification[J]. Magn Reson Med, 2022, 87(4): 2044-2052. DOI: 10.1002/mrm.29074.
[23]
She HJ, Greer JS, Zhang S, et al. Accelerating chemical exchange saturation transfer MRI with parallel blind compressed sensing[J]. Magn Reson Med, 2019, 81(1): 504-513. DOI: 10.1002/mrm.27400.
[24]
Li YR, Xie DF, Cember A, et al. Accelerating GluCEST imaging using deep learning for B0 correction[J]. Magn Reson Med, 2020, 84(4): 1724-1733. DOI: 10.1002/mrm.28289.
[25]
Guo CL, Wu J, Rosenberg JT, et al. Fast chemical exchange saturation transfer imaging based on PROPELLER acquisition and deep neural network reconstruction[J]. Magn Reson Med, 2020, 84(6): 3192-3205. DOI: 10.1002/mrm.28376.
[26]
Kang B, Kim B, Park H, et al. Learning-based optimization of acquisition schedule for magnetization transfer contrast MR fingerprinting[J/OL]. 2021 [2022-09-12]. https://doi.org/10.1002/nbm.4662. DOI: 10.1002/nbm.4662.
[27]
Bie CX, Li YG, Zhou Y, et al. Deep learning-based classification of preclinical breast cancer tumor models using chemical exchange saturation transfer magnetic resonance imaging[J/OL]. NMR Biomed, 2022 [2022-09-12]. https://doi.org/10.1002/nbm.4626. DOI: 10.1002/nbm.4626.
[28]
Song XL, Xu JD, Xia SL, et al. Multi-echo length and offset VARied saturation (MeLOVARS) method for improved CEST imaging[J]. Magn Reson Med, 2015, 73(2): 488-496. DOI: 10.1002/mrm.25567.
[29]
Zhang LH, Liang YH, Chen YR, et al. Enhanced CEST MRI using the residual of inversed Z-spectra for ischemia detection[J]. IEEE Access, 8: 147323-147336. DOI: 10.1109/ACCESS.2020.3014339.
[30]
Zhang XY, Wang F, Li H, et al. Accuracy in the quantification of chemical exchange saturation transfer (CEST) and relayed nuclear Overhauser enhancement (rNOE) saturation transfer effects[J/OL]. NMR Biomed, 2017 [2022-09-12]. https://doi.org/10.1002/nbm.3716. DOI: 10.1002/nbm.3716.
[31]
Zhang XY, Wang F, Afzal A, et al. A new NOE-mediated MT signal at around-1.6 ppm for detecting ischemic stroke in rat brain[J]. Magn Reson Imaging, 2016, 34(8): 1100-1106. DOI: 10.1016/j.mri.2016.05.002.
[32]
Zaiss M, Schmitt B, Bachert P. Quantitative separation of CEST effect from magnetization transfer and spillover effects by Lorentzian-line-fit analysis of z-spectra[J]. J Magn Reson, 2011, 211(2): 149-155. DOI: 10.1016/j.jmr.2011.05.001.
[33]
Liu YX, Li J, Ji HF, et al. Comparisons of glutamate in the brains of Alzheimer's disease mice under chemical exchange saturation transfer imaging based on machine learning analysis[J/OL]. Front Neurosci, 2022 [2022-09-12]. https://doi.org/10.3389/fnins.2022.838157. DOI: 10.3389/fnins.2022.838157.
[34]
Glang F, Fabian MS, German A, et al. Linear projection-based chemical exchange saturation transfer parameter estimation[J/OL]. NMR Biomed, 2022 [2022-09-12]. https://doi.org/10.1002/nbm.4697. DOI: 10.1002/nbm.4697.
[35]
Zaiss M, Schuppert M, Deshmane A, et al. Chemical exchange saturation transfer MRI contrast in the human brain at 9.4 T[J]. Neuroimage, 2018, 179: 144-155. DOI: 10.1016/j.neuroimage.2018.06.026.
[36]
Heo HY, Lee DH, Zhang Y, et al. Insight into the quantitative metrics of chemical exchange saturation transfer (CEST) imaging[J]. Magn Reson Med, 2017, 77(5): 1853-1865. DOI: 10.1002/mrm.26264.
[37]
Dou WQ, Lin CE, Ding HY, et al. Chemical exchange saturation transfer magnetic resonance imaging and its main and potential applications in pre-clinical and clinical studies[J]. Quant Imaging Med Surg, 2019, 9(10): 1747-1766. DOI: 10.21037/qims.2019.10.03.
[38]
Chen L, Wei ZL, Chan KWY, et al. Protein aggregation linked to Alzheimer's disease revealed by saturation transfer MRI[J]. Neuroimage, 2019, 188: 380-390. DOI: 10.1016/j.neuroimage.2018.12.018.
[39]
Jin T, Mehrens H, Wang P, et al. Chemical exchange-sensitive spin-lock MRI of glucose analog 3-O-methyl-d-glucose in normal and ischemic brain[J]. J Cereb Blood Flow Metab, 2018, 38(5): 869-880. DOI: 10.1177/0271678x17707419.
[40]
Kujawa A, Kim M, Demetriou E, et al. Assessment of a clinically feasible Bayesian fitting algorithm using a simplified description of Chemical Exchange Saturation Transfer (CEST) imaging[J]. J Magn Reson, 2019, 300: 120-134. DOI: 10.1016/j.jmr.2019.01.006.
[41]
Mehrabian H, Myrehaug S, Soliman H, et al. Evaluation of glioblastoma response to therapy with chemical exchange saturation transfer[J]. Int J Radiat Oncol Biol Phys, 2018, 101(3): 713-723. DOI: 10.1016/j.ijrobp.2018.03.057.
[42]
Ryoo D, Xu X, Li YG, et al. Detection and quantification of hydrogen peroxide in aqueous solutions using chemical exchange saturation transfer[J]. Anal Chem, 2017, 89(14): 7758-7764. DOI: 10.1021/acs.analchem.7b01763.
[43]
Akbari H, Kazerooni AF, Ware JB, et al. Quantification of tumor microenvironment acidity in glioblastoma using principal component analysis of dynamic susceptibility contrast enhanced MR imaging[J/OL]. Sci Rep, 2021 [2022-09-12]. https://doi.org/10.1038/s41598-021-94560-3. DOI: 10.1038/s41598-021-94560-3.
[44]
Cai KJ, Singh A, Poptani H, et al. CEST signal at 2ppm (CEST@2ppm) from Z-spectral fitting correlates with creatine distribution in brain tumor[J]. NMR Biomed, 2015, 28(1): 1-8. DOI: 10.1002/nbm.3216.
[45]
Cai KJ, Haris M, Singh A, et al. Magnetic resonance imaging of glutamate[J]. Nat Med, 2012, 18(2): 302-306. DOI: 10.1038/nm.2615.
[46]
Glang F, Deshmane A, Prokudin S, et al. DeepCEST 3T: robust MRI parameter determination and uncertainty quantification with neural networks-application to CEST imaging of the human brain at 3T[J]. Magn Reson Med, 2020, 84(1): 450-466. DOI: 10.1002/mrm.28117.
[47]
Huang JP, Lai JHC, Tse KH, et al. Deep neural network based CEST and AREX processing: application in imaging a model of Alzheimer's disease at 3 T[J]. Magn Reson Med, 2022, 87(3): 1529-1545. DOI: 10.1002/mrm.29044.
[48]
Zhao YC, Chen YB, Chen YR, et al. A fully convolutional network (FCN) based automated ischemic stroke segment method using chemical exchange saturation transfer imaging[J]. Med Phys, 2022, 49(3): 1635-1647. DOI: 10.1002/mp.15483.
[49]
Guo PF, Unberath M, Heo HY, et al. Learning-based analysis of amide proton transfer-weighted MRI to identify true progression in glioma patients[J/OL]. Neuroimage Clin, 2022 [2022-09-12]. https://doi.org/10.1016/j.nicl.2022.103121. DOI: 10.1016/j.nicl.2022.103121.
[50]
Zaiss M, Deshmane A, Schuppert M, et al. DeepCEST: 9.4 T Chemical exchange saturation transfer MRI contrast predicted from 3 T data - a proof of concept study[J]. Magn Reson Med, 2019, 81(6): 3901-3914. DOI: 10.1002/mrm.27690.
[51]
Perlman O, Farrar CT, Heo HY. MR fingerprinting for semisolid magnetization transfer and chemical exchange saturation transfer quantification[J/OL]. NMR Biomed, 2022 [2022-09-12]. https://doi.org/10.1002/nbm.4710. DOI: 10.1002/nbm.4710.
[52]
Kim B, Schär M, Park H, et al. A deep learning approach for magnetization transfer contrast MR fingerprinting and chemical exchange saturation transfer imaging[J/OL]. NeuroImage, 2020 [2022-09-12]. https://doi.org/10.1016/j.neuroimage.2020.117165. DOI: 10.1016/j.neuroimage.2020.117165.
[53]
Zhou ZW, Han P, Zhou B, et al. Chemical exchange saturation transfer fingerprinting for exchange rate quantification[J]. Magn Reson Med, 2018, 80(4): 1352-1363. DOI: 10.1002/mrm.27363.
[54]
Kang B, Kim B, Schär M, et al. Unsupervised learning for magnetization transfer contrast MR fingerprinting: application to CEST and nuclear Overhauser enhancement imaging[J]. Magn Reson Med, 2021, 85(4): 2040-2054. DOI: 10.1002/mrm.28573.
[55]
Karunanithy G, Yuwen TR, Kay LE, et al. Towards autonomous analysis of chemical exchange saturation transfer experiments using deep neural networks[J]. J Biomol NMR, 2022, 76(3): 75-86. DOI: 10.1007/s10858-022-00395-z.
[56]
Chen L, Schär M, Chan KWY, et al. In vivo imaging of phosphocreatine with artificial neural networks[J/OL]. Nat Commun, 2020 [2022-09-12]. https://doi.org/10.1038/s41467-020-14874-0. DOI: 10.1038/s41467-020-14874-0.

上一篇 中医康复疗法干预膝骨关节炎的脑网络机制研究进展
下一篇 一种高分辨AQP4分子磁共振成像技术有望无创预测胶质瘤治疗敏感性
  
诚聘英才 | 广告合作 | 免责声明 | 版权声明
联系电话:010-67113815
京ICP备19028836号-2