搜索

x

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于深度神经网络的时空编码磁共振成像超分辨率重建方法

向鹏程 蔡聪波 王杰超 蔡淑惠 陈忠

引用本文:
Citation:

基于深度神经网络的时空编码磁共振成像超分辨率重建方法

向鹏程, 蔡聪波, 王杰超, 蔡淑惠, 陈忠
cstr: 32037.14.aps.71.20211754

Super-resolved reconstruction method for spatiotemporally encoded magnetic resonance imaging based on deep neural network

Xiang Peng-Cheng, Cai Cong-Bo, Wang Jie-Chao, Cai Shu-Hui, Chen Zhong
cstr: 32037.14.aps.71.20211754
Article Text (iFLYTEK Translation)
PDF
HTML
导出引用
  • 单扫描时空编码磁共振成像是一种新型超快速磁共振成像技术, 它对磁场不均匀和化学位移伪影有较强的抵抗性, 但是其固有的空间分辨率较低, 因此通常需要进行超分辨率重建, 以在不增加采样点数的情况下提高时空编码磁共振图像的空间分辨率. 然而, 现有的重建方法存在迭代求解时间长、重建结果有混叠伪影残留等问题. 为此, 本文提出了一种基于深度神经网络的单扫描时空编码磁共振成像超分辨率重建方法. 该方法采用模拟样本训练深度神经网络, 再利用训练好的网络模型对实际采样信号进行重建. 数值模拟、水模和活体鼠脑的实验结果表明, 该方法能快速重建出无残留混叠伪影、纹理信息清楚的超分辨率时空编码磁共振图像. 适当增加训练样本数量以及在训练样本中加入适当的随机噪声水平, 有助于改善重建效果.
    Single-shot spatiotemporally-encoded magnetic resonance imaging (SPEN MRI) is a novel ultrafast MRI technology. The SPEN MRI possesses great resistance to inhomogeneous B0 magnetic field and chemical shift effect. However, it has inherently low spatial resolution, and the super-resolved reconstruction is required to improve the spatial resolution of SPEN MRI image without additional signal acquisition. Several super-resolved reconstruction methods have been proposed, but they all suffer the problems of long iterative solution time and/or aliasing artifacts residue in the reconstructed results. In this paper, a super-resolved reconstruction method is proposed for single-shot SPEN MRI based on deep neural network. In this method the simulation samples are used to train the deep neural network, and then the trained network model is adopted to reconstruct the real sampled signals. Experimental results of numerical simulation, water phantom and in vivo rat brain show that this method can quickly reconstruct a super-resolved SPEN image with no residual aliasing artifacts, and clear texture information. An appropriate number of training samples and an appropriate random noise level for training samples contribute to improving the reconstruction results.
      通信作者: 蔡淑惠, shcai@xmu.edu.cn
    • 基金项目: 国家自然科学基金(批准号: 11775184, 82071913, U1805261)和福建省科技计划项目(批准号: 2019Y0001)资助的课题
      Corresponding author: Cai Shu-Hui, shcai@xmu.edu.cn
    • Funds: Project supported by the National Natural Science Foundation of China (Grant Nos. 11775184, 82071913, U1805261) and the Science and Technology Project of Fujian Province of China (Grant No. 2019Y0001)
    [1]

    Shrot Y, Frydman L 2005 J. Magn. Reson. 172 179Google Scholar

    [2]

    Tal A, Frydman L 2006 J. Magn. Reson. 182 179Google Scholar

    [3]

    Solomon E, Avni R, Hadas R, Raz T, Garbow J R, Bendel P, Frydman L, Neeman M 2014 Proc. Nati. Acad. Sci. USA 111 10353Google Scholar

    [4]

    Ben-Eliezer N, Irani M, Frydman L 2010 Magn. Reson. Med. 63 1594Google Scholar

    [5]

    Chen Y, Li J, Qu X B, Chen L, Cai C B, Cai S H, Zhong J H, Chen Z 2013 Magn. Reson. Med. 69 1326Google Scholar

    [6]

    Cai C B, Dong J Y, Cai S H, Li J, Chen Y, Bao L J, Chen Z 2013 J. Magn. Reson. 228 136Google Scholar

    [7]

    Chen L, Li J, Zhang M, Cai S H, Zhang T, Cai C B, Chen Z 2015 Med. Image Anal. 23 1Google Scholar

    [8]

    Aliotta E, Nourzadeh H, Sanders J, Muller D, Ennis D B 2019 Med. Phys. 46 1581Google Scholar

    [9]

    Chun J, Zhang H, Gach H M, Olberg S, Mazur T, Green O, Kim T, Kim H, Kim J S, Mutic S, Park J C 2019 Med. Phys. 46 4148Google Scholar

    [10]

    Le M H, Chen J, Wang L, Wang Z, Liu W, Cheng K T, Yang X 2017 Phys. Med. Biol. 62 6497Google Scholar

    [11]

    Liu Y, Lei Y, Wang Y, Wang T, Ren L, Lin L, McDonald M, Curran W J, Liu T, Zhou J, Yang X 2019 Phys. Med. Biol. 64 145015Google Scholar

    [12]

    罗伶俐, 王远军 2020 中国医学物理学杂志 37 873Google Scholar

    Luo L L, Wang Y J 2020 Chin. J. Med. Phys. 37 873Google Scholar

    [13]

    王甜甜, 王慧, 朱艳春, 王丽嘉 2021 物理学报 70 228701Google Scholar

    Wang T T, Wang H, Zhu Y C, Wang L J 2021 Acta Phys. Sin. 70 228701Google Scholar

    [14]

    Schlemper J, Caballero J, Hajnal J V, Price A N, Rueckert D 2018 IEEE Trans. Med. Imaging 37 491Google Scholar

    [15]

    Shi J, Liu Q, Wang C, Zhang Q, Ying S, Xu H 2018 Phys. Med. Biol. 63 085011Google Scholar

    [16]

    Quan T M, Nguyen-Duc T, Jeong W K 2018 IEEE Trans. Med. Imaging 37 1488Google Scholar

    [17]

    Guo C L, Wu J, Rosenberg J T, Roussel T, Cai S H, Cai C B 2020 Magn. Reson. Med. 84 3192Google Scholar

    [18]

    Akkus Z, Galimzianova A, Hoogi A, Rubin D L, Erickson B J 2017 J. Digit. Imaging 30 449Google Scholar

    [19]

    Zhang J, Wu J, Chen S J, Zhang Z Y, Cai S H, Cai C B, Chen Z 2019 IEEE Trans. Med. Imaging 38 1801Google Scholar

    [20]

    Ronneberger O, Fischer P, Brox T 2015 18th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI) Munich, Germany, October 5–9, 2015 p234

    [21]

    Siddique N, Paheding S, Elkin C P, Devabhaktuni V 2021 IEEE Access 9 82031Google Scholar

    [22]

    Yang G, Yu S M, Dong H, Slabaugh G, Dragotti P L, Ye X J, Liu F D, Arridge S, Keegan J, Guo Y K, Firmin D 2018 IEEE Trans. Med. Imaging 37 1310Google Scholar

    [23]

    Liu F, Velikina J V, Block W F, Kijowski R, Samsonov A A 2017 IEEE Trans. Med. Imaging 36 527Google Scholar

  • 图 1  本文所用SPEN MRI脉冲序列

    Fig. 1.  Pulse sequence of SPEN MRI used in this study.

    图 2  U-Net网络结构示意图

    Fig. 2.  Diagram of U-Net network structure.

    图 3  采用模拟样本训练U-Net的训练误差曲线和验证误差曲线

    Fig. 3.  The training error curve and validation error curve of U-Net trained with simulated samples.

    图 4  数值模拟结果, 红色矩形所围区域放大显示于相应图的右下角

    Fig. 4.  Numerical simulation results. The region enclosed by the red rectangle is enlarged and displayed in the lower right corner of the corresponding figure.

    图 5  不同成像序列和重建方法得到的水模图像, 红色矩形所围区域放大显示于相应图的左下角

    Fig. 5.  Water phantom images obtained by different imaging sequences and reconstruction methods. The region enclosed by the red rectangle is enlarged and displayed in the lower left corner of the corresponding figure.

    图 6  不同成像序列和重建方法得到的活体鼠脑图像

    Fig. 6.  In vivo rat brain images obtained by different imaging sequences and reconstruction methods.

    图 7  采用不同数量训练样本训练的5层U-Net网络重建的图像

    Fig. 7.  Images reconstructed by 5-layer U-Net trained with different amount of training samples.

    图 8  采用不同噪声水平样本训练的5层U-Net网络重建的水模图像

    Fig. 8.  Water phantom images reconstructed by 5-layer U-Net trained with samples having different noise levels.

  • [1]

    Shrot Y, Frydman L 2005 J. Magn. Reson. 172 179Google Scholar

    [2]

    Tal A, Frydman L 2006 J. Magn. Reson. 182 179Google Scholar

    [3]

    Solomon E, Avni R, Hadas R, Raz T, Garbow J R, Bendel P, Frydman L, Neeman M 2014 Proc. Nati. Acad. Sci. USA 111 10353Google Scholar

    [4]

    Ben-Eliezer N, Irani M, Frydman L 2010 Magn. Reson. Med. 63 1594Google Scholar

    [5]

    Chen Y, Li J, Qu X B, Chen L, Cai C B, Cai S H, Zhong J H, Chen Z 2013 Magn. Reson. Med. 69 1326Google Scholar

    [6]

    Cai C B, Dong J Y, Cai S H, Li J, Chen Y, Bao L J, Chen Z 2013 J. Magn. Reson. 228 136Google Scholar

    [7]

    Chen L, Li J, Zhang M, Cai S H, Zhang T, Cai C B, Chen Z 2015 Med. Image Anal. 23 1Google Scholar

    [8]

    Aliotta E, Nourzadeh H, Sanders J, Muller D, Ennis D B 2019 Med. Phys. 46 1581Google Scholar

    [9]

    Chun J, Zhang H, Gach H M, Olberg S, Mazur T, Green O, Kim T, Kim H, Kim J S, Mutic S, Park J C 2019 Med. Phys. 46 4148Google Scholar

    [10]

    Le M H, Chen J, Wang L, Wang Z, Liu W, Cheng K T, Yang X 2017 Phys. Med. Biol. 62 6497Google Scholar

    [11]

    Liu Y, Lei Y, Wang Y, Wang T, Ren L, Lin L, McDonald M, Curran W J, Liu T, Zhou J, Yang X 2019 Phys. Med. Biol. 64 145015Google Scholar

    [12]

    罗伶俐, 王远军 2020 中国医学物理学杂志 37 873Google Scholar

    Luo L L, Wang Y J 2020 Chin. J. Med. Phys. 37 873Google Scholar

    [13]

    王甜甜, 王慧, 朱艳春, 王丽嘉 2021 物理学报 70 228701Google Scholar

    Wang T T, Wang H, Zhu Y C, Wang L J 2021 Acta Phys. Sin. 70 228701Google Scholar

    [14]

    Schlemper J, Caballero J, Hajnal J V, Price A N, Rueckert D 2018 IEEE Trans. Med. Imaging 37 491Google Scholar

    [15]

    Shi J, Liu Q, Wang C, Zhang Q, Ying S, Xu H 2018 Phys. Med. Biol. 63 085011Google Scholar

    [16]

    Quan T M, Nguyen-Duc T, Jeong W K 2018 IEEE Trans. Med. Imaging 37 1488Google Scholar

    [17]

    Guo C L, Wu J, Rosenberg J T, Roussel T, Cai S H, Cai C B 2020 Magn. Reson. Med. 84 3192Google Scholar

    [18]

    Akkus Z, Galimzianova A, Hoogi A, Rubin D L, Erickson B J 2017 J. Digit. Imaging 30 449Google Scholar

    [19]

    Zhang J, Wu J, Chen S J, Zhang Z Y, Cai S H, Cai C B, Chen Z 2019 IEEE Trans. Med. Imaging 38 1801Google Scholar

    [20]

    Ronneberger O, Fischer P, Brox T 2015 18th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI) Munich, Germany, October 5–9, 2015 p234

    [21]

    Siddique N, Paheding S, Elkin C P, Devabhaktuni V 2021 IEEE Access 9 82031Google Scholar

    [22]

    Yang G, Yu S M, Dong H, Slabaugh G, Dragotti P L, Ye X J, Liu F D, Arridge S, Keegan J, Guo Y K, Firmin D 2018 IEEE Trans. Med. Imaging 37 1310Google Scholar

    [23]

    Liu F, Velikina J V, Block W F, Kijowski R, Samsonov A A 2017 IEEE Trans. Med. Imaging 36 527Google Scholar

计量
  • 文章访问数:  7017
  • PDF下载量:  153
  • 被引次数: 0
出版历程
  • 收稿日期:  2021-09-21
  • 修回日期:  2021-11-10
  • 上网日期:  2022-03-01
  • 刊出日期:  2022-03-05

/

返回文章
返回