基于Deep CG的肺部电阻抗成像方法

doi:10.3969/j.issn.0258-8021.2023.02.003

摘要
图/表
参考文献
相关文章 (15)

全文: PDF (7143 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要针对电阻抗图像重建空间分辨率不足问题,基于深度学习理论提出一种共轭梯度快速预重建与深度堆栈式自编码器后处理的电阻抗成像方法(Deep CG)。该方法的核心思想是:融合数值重建算法与深度学习算法,使胸腔内肺部的结构和电导率分布更加精准。首先采用共轭梯度算法进行图像预重建,获得边界电压与胸腔内部电导率分布的预映射关系;再采用深度堆栈式自编码器,将编码和解码层级连接,充分利用不同空间特征信息,实现特征提取和图像重建;最后根据公开的80名临床患者的CT结构图像构建了数据集,采用混合式监督训练方法调参,不仅避免了深度网络中信息流和梯度流弥散问题,而且优化了算法模型。采用图像相对误差、相关系数进行量化指标评价,并与常用的数值图像重建算法和全连接神经网络模型进行对比。结果显示,Deep CG算法的比常用图像重建算法图像和全连接神经网络模型相对误差从0.50和0.24降低到0.11,相关系数由0.80和0.90提高到0.96。该方法获得了空间分辨率高,尤其边界更清晰的电阻抗图像,有望进一步推动EIT技术在临床的应用研究。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	王子辰
	付荣
	张新宇
	王迪
	陈晓艳

关键词 ：电阻抗成像, 堆栈式自编码器, 深度学习, 图像重建, 共轭梯度

Abstract：To improve the spatial resolution of electrical impedance tomography, a novel method based on conjugate gradient (CG) with rapid pre-reconstruction and deep stack autoencoder post-processing was proposed (Deep CG). The core idea is that the merging ofnumerical reconstruction algorithm with deeplearning-based method makes the structure and conductivity distribution of the thorax more accurate. Firstly, the mathematical reconstruction algorithm CG was adopted to pre-reconstruct the coarse image, and the mapping between boundary voltage and conductivity distribution in the chest was achieved. Next, to take full advantages of the different spatial features, the stack autoencoder was employed to connect the encoding and the decoding modules hierarchically, which realized the feature extraction (FE) and the image reconstruction (IR). Finally, to train the model, the dataset was constructed from the number of 400 clinical CT slices, a mixed supervised method was employed to adjust the model parameters, which not only avoided the dispersion of the information flow and gradient flow, but alsooptimized the parameters of Deep CG method. The relative error (RE) and correlation coefficient (CC) were adopted to evaluate the image quantitively. The images were compared to the traditionalnumerical algorithm and a full connected neural network. The results showed that the RE was decreased to 0.11 from 0.5 and 0.24, and the CC was improved to 0.96 from 0.8 and 0.9. The proposed method was able to reconstruct EIT images with higher spatial resolution and clear boundary,which is expected to put forward EIT techniques to the further applications and researches.

Key words： electrical impedance imaging stack autoencoder deep learning image reconstruction conjugate gradient

收稿日期: 2021-04-26

PACS:

R318

通讯作者: ^*E-mail: cxywxr@tust.edu.cn

引用本文:

王子辰, 付荣, 张新宇, 王迪, 陈晓艳. 基于Deep CG的肺部电阻抗成像方法[J]. 中国生物医学工程学报, 2023, 42(2): 148-157.
Wang Zichen, Fu Rong, Zhang Xinyu, Wang Di, Chen Xiaoyan. Reconstruction of Thorax Image Based on Deep CG Method for Electrical Impedance Tomography. Chinese Journal of Biomedical Engineering, 2023, 42(2): 148-157.

链接本文:

http://cjbme.csbme.org/CN/10.3969/j.issn.0258-8021.2023.02.003 或 http://cjbme.csbme.org/CN/Y2023/V42/I2/148

[1] 王琦,陈晓静,汪剑鸣,等. 基于人体结构先验信息的胸部电阻抗成像方法[J]. 中国生物医学工程学报,2019,38(1):35-43.
[2] 董秀珍. 生物电阻抗成像研究的现状与挑战[J]. 中国生物医学工程学报,2008,27(5):641-643.
[3] Bayford,RH. Bioimpedance tomography (electrical impedance tomography)[J]. Annual Review of Biomedical Engineering, 2006, 8(1):63-91.
[4] Brown,BH. Electrical impedance tomography (EIT):a review[J]. Journal of Medical Engineering & Technology, 2003, 27(3):97-108.
[5] 任尚杰,董峰,谭超. 基于边界元与有限元耦合的ERT图像重建算法[J]. 工程热物理学报,2014(5):902-905.
[6] Huang Y,Chen M,He W,et al. A parameter selecting method of newton′s one-step error reconstructor in electrical impedance tomography imaging[J]. Journal of Information & Computational Science,2013,10(3):729-741.
[7] Yang Y,Wu H,Jia J. Image reconstruction for electrical impedance tomography using enhanced adaptive group sparsity with total variation[J]. IEEE Sensors Journal,2017,17(17):5589-5598.
[8] Yang Y,Jia J. An image reconstruction algorithm for electrical impedance tomography using adaptive group sparsity constraint[J]. IEEE Transactions on Instrumentation & Measurement,2017,66(9):2295-2305.
[9] Hinton GE,Salakhutdinov RR. Reducing the dimensionality of data with neural networks[J]. Science(New York),2006,313(5786):504-507.
[10] Martin S,Choi C. On the influence of spread constant in radial basis networks for electrical impedance tomography[J]. Physiological Measurement,2016,37(6):801-819.
[11] Borsoi RA,Aya JCC,Costa GH,et al. Super-resolution reconstruction of electrical impedance tomography images[J]. Computers and Electrical Engineering,2018,69:1-13.
[12] Huang SW,Cheng HM and Lin SF. Improved imaging resolution of electrical impedance tomography using artificial neural networks for image reconstruction[C]//The 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). Berlin:IEEE, 2019: 1551-1554.
[13] Zheng J,Peng L. An autoencoder-based image reconstruction for electrical capacitance tomography[J]. IEEE Sensors Journal,2018,18(13):5464-5474.
[14] Li X,Zhou Y,Wang J,et al. A novel deep neural network method for electrical impedance tomography[J]. Transactions of the Institute of Measurement and Control,2019,41(14):4035-4049.
[15] 李峰,谭超,董峰. 全连接深度网络的电学层析成像算法[J]. 工程热物理学报,2019(7):1526-1531.
[16] Tan C,Lv S,Dong F,et al. Image reconstruction based on convolutional neural network for electrical resistance tomography[J]. IEEE Sensors Journal,2018,19(1):196-204.
[17] Li F,Tan C,Dong F,et al. V-Net deep imaging method for electrical resistance tomography[J]. IEEE Sensors Journal,2020,20(12):6460-6469.
[18] Li F,Tan C,Dong F. Electrical resistance tomography image reconstruction with densely connected convolutional neural network[J]. IEEE Transactions on Instrumentation and Measurement,2020,70:1-11.
[19] Wu Y,Chen B,Liu K,et al. Shape reconstruction with multiphase conductivity for electrical impedance tomography using improved convolutional neural network method[J]. IEEE Sensors Journal,2021,21(7):9277-9287.
[20] Jin KS,Kang CK,Jargal A,et al. A learning-based method for solving ill-posed nonlinear inverse problems:a simulation study of lung EIT[J]. SIAM Journal on Imaging Sciences,2019,12(3):1275-1295.
[21] Ren S,Sun K,Tan C,et al. A two-stage deep learning method for robust shape reconstruction with electrical impedance tomography[J]. IEEE Transactions on Instrumentation and Measurement,2019,69(7):4887-4897.
[22] Hamilton SJ,Hauptmann A. Deep D-bar:real time electrical impedance tomography imaging with deep neural networks[J]. IEEE Transactions on Medical Imaging,2017,37(10):2367-2377.
[23] Capps M and Mueller JL. Reconstruction of organ boundaries with deep learning in the D-Bar method for electrical impedance tomography[J]. IEEE Transactions on Biomedical Engineering,2021,68(3):826-833.
[24] Hestenes MR,Stiefel EL. Methods of conjugate gradients for solving linear systems[J]. Journal of Research of the National Bureau of Standards (United States),1952,49(6):409-435.
[25] Clark K,Vendt B,Smith K,et al. The cancer imaging archive (TCIA):maintaining and operating a public information repository[J]. Journal of Digital Imaging,2013,26(6):1045-1057.
[26] Gelidi SD,Seifnaraghi N,Bardill A,et al. Towards a thoracic conductive phantom for EIT[J]. Medical Engineering & Physics,2020,77:88-94.
[27] Li X,Chen X,Wang Q,et al. Electrical-impedance-tomography imaging based on a new three-dimensional thorax model for assessing the extent of lung injury[J]. AIP Advances,2019,9(12):125310.
[28] Wang H,Lei T,Zhang C. An image reconstruction algorithm based on total variation with adaptive mesh refinement for ECT[J]. Flow Measurement and Instrumentation,2007,18(5-6):262-267.
[29] Xu Y,Pei Y,Dong F. An adaptive Tikhonov regularization parameter choice method for electrical resistance tomography[J]. Flow Measurement and Instrumentation,2016,50:1-12.
[30] Wang H,Zhu X,Zhang L. Conjugate Gradient Algorithm for Electrical Capacitance Tomography[J]. Journal of Tianjin University,2005,38(1):1-4.