《计算机应用》唯一官方网站 ›› 2025, Vol. 45 ›› Issue (10): 3381-3389.DOI: 10.11772/j.issn.1001-9081.2024101501

• 前沿与综合应用 • 上一篇    

基于残差复卷积网络的7T超高场磁共振并行成像算法

高照耀1,2, 张展2(), 胡亮亮3, 许光宇1, 周胜4, 胡雨欣1,2, 林子捷5, 周超2,5   

  1. 1.安徽理工大学 计算机科学与工程学院,安徽 淮南 232001
    2.合肥综合性国家科学中心能源研究院(安徽省能源实验室) 超导技术应用研究中心,合肥 230031
    3.合肥工业大学 仪器科学与光电工程学院,合肥 230009
    4.合肥曦合超导科技有限公司,合肥 230031
    5.中国科学院 等离子体物理研究所,合肥 230031
  • 收稿日期:2024-10-24 修回日期:2025-01-10 接受日期:2025-01-16 发布日期:2025-02-07 出版日期:2025-10-10
  • 通讯作者: 张展
  • 作者简介:高照耀(1999—),男,安徽滁州人,硕士研究生,主要研究方向:信号处理、核磁共振图像重建
    张展(1986—),男,河北邢台人,副研究员,博士,主要研究方向:超导磁体、超导核磁共振系统 Email:zhanzhang@ie.ah.cn
    胡亮亮(1981—),男,安徽岳西人,讲师,博士,主要研究方向:核磁共振仪器、磁共振弹性成像
    许光宇(1976—),男,安徽合肥人,副教授,博士,CCF会员,主要研究方向:数字图像处理、机器学习。
  • 基金资助:
    合肥综合性国家科学中心能源研究院(安徽省能源实验室)项目(24JYCD01);安徽高校协同创新项目(GXXT-2022-011)

7T ultra-high field magnetic resonance parallel imaging algorithm based on residual complex convolution network

Zhaoyao GAO1,2, Zhan ZHANG2(), Liangliang HU3, Guangyu XU1, Sheng ZHOU4, Yuxin HU1,2, Zijie LIN5, Chao ZHOU2,5   

  1. 1.School of Computer Science and Engineering,Anhui University of Science and Technology,Huainan Anhui 232001,China
    2.Center for Applied Research of Superconducting Technology,Institute of Energy,Hefei Comprehensive National Science Center (Anhui Energy Laboratory),Hefei Anhui 230031,China
    3.School of Instrument Science and Opto-electronics Engineering,Hefei University of Technology,Hefei Anhui 230009,China
    4.Hefei Xihe Superconducting Technology Company Limited,Hefei Anhui 230031,China
    5.Institute of Plasma Physics,Chinese Academy of Sciences,Hefei Anhui 230031,China
  • Received:2024-10-24 Revised:2025-01-10 Accepted:2025-01-16 Online:2025-02-07 Published:2025-10-10
  • Contact: Zhan ZHANG
  • About author:GAO Zhaoyao, born in 1999, M. S. candidate. His research interests include signal processing, magnetic resonance imaging reconstruction.
    ZHANG Zhan, born in 1986, Ph. D., associate research fellow. His research interests include superconducting magnet, superconducting magnetic resonance systems.
    HU Liangliang, born in 1981, Ph. D., lecturer. His research interests include magnetic resonance instruments, magnetic resonance elastography.
    XU Guangyu, born in 1976, Ph. D., associate professor. His research interests include digital image processing, machine learning.
  • Supported by:
    Research Project of Hefei Comprehensive National Science Center (Anhui Energy Laboratory)(24JYCD01);Collaborative Innovation Project of Anhui Universities(GXXT-2022-011)

摘要:

并行成像技术可以帮助解决超高场强磁共振成像(MRI)中的射频能量沉积、图像均匀性的问题,缩短扫描时间,减少运动伪影,并提升数据采集速度。为了提高对MRI复值数据的特征提取能力,减少并行成像欠采样所引起的卷褶伪影,提出基于K空间插值的残差复卷积鲁棒人工神经网络(RCRAKI)。所提算法将原始欠采样MRI扫描数据作为输入,利用残差结构结合线性与非线性重建方法的优势,在残差连接部分利用卷积创建线性重建基线,主路径利用多层复卷积补偿基线缺陷,最终重建出伪影更少的磁共振(MR)图像。在合肥综合性国家科学中心能源研究院自主研发的7T超高场磁共振设备采集的数据上进行实验,并将RCRAKI与基于K空间插值的残差鲁棒人工神经网络(rRAKI)在自动校准信号(ACS)数为40、加速比为8的采样率下进行小鼠不同解剖切面成像质量对比。实验结果表明:在矢状位下,所提算法的标准化均方根误差(NRMSE)指标下降了59.74%,结构相似度(SSIM)指标提升了0.45%,峰值信噪比(PSNR)指标提升了13.04%;在横断位下,所提算法的NRMSE指标降低了7.97%,SSIM指标略有改善(提高了0.005%),PSNR指标提升了1.09%;在冠状位下,所提算法的NRMSE指标下降了35.03%,PSNR指标提升了5.60%,SSIM指标提升了0.98%。可见,RCRAKI在不同解剖切面的MRI数据上均表现出良好的性能,在高加速比采样率下能够减小噪声放大的影响,并重建出细节更清晰的MR图像。

关键词: 超高场磁共振, 并行成像, 欠采样图像, 复卷积, 深度学习

Abstract:

Parallel imaging techniques can help solving problems of radiofrequency energy deposition and image inhomogeneity, reducing scan time, lowering motion artifacts, and accelerating data acquisition in ultra-high field Magnetic Resonance Imaging (MRI). To enhance feature extraction ability to MRI complex-valued data and reduce wrap-around artifacts caused by under-sampling in parallel imaging, a Residual Complex convolution scan-specific Robust Artificial-neural-networks for K-space Interpolation (RCRAKI) was proposed. In the algorithm, the raw under-sampled MRI scan data was taken as input, and the advantages of both linear and nonlinear reconstruction methods were combined with a residual structure. In the residual connection part, convolution was used to create a linear reconstruction baseline, while multiple layers of complex convolution were utilized in the main path to compensate for baseline defects, ultimately reconstructing Magnetic Resonance (MR) images with fewer artifacts. Experiments were conducted on data acquired from a 7T ultra-high field MR device developed by the Institute of Energy of Hefei Comprehensive National Science Center, and RCRAKI was compared with residual scan-specific Robust Artificial-neural-networks for K-space Interpolation (rRAKI) under a sampling rate of 40 Automatic Calibration Signals (ACSs) and 8 speedup ratio for mouse imaging quality across different anatomical planes. Experimental results show that in sagittal plane, the proposed algorithm has the Normalized Root Mean Squared Error (NRMSE) decreased by 59.74%, the Structural SIMilarity (SSIM) increased by 0.45%, and the Peak Signal-to-Noise Ratio (PSNR) increased by 13.04%; in axial plane, the proposed algorithm has the NRMSE decreased by 7.97%, the SSIM improved slightly (by 0.005%), and the PSNR increased by 1.09%; in coronal plane, the proposed algorithm has the NRMSE decreased by 35.03%, the PSNR increased by 5.60%, and the SSIM increased by 0.98%. It can be seen that RCRAKI performs well on all the different anatomical planes of MRI data, can reduce the influence of noise amplification at high speedup ratio, and reconstruct MR images with clearer details.

Key words: ultra-high field Magnetic Resonance (MR), parallel imaging, under-sampled image, complex convolution, deep learning

中图分类号: