A Regularization-Guided Deep Imaging Method for Electrical Impedance Tomography

Fu, Rong; Wang, Zichen; Zhang, Xinyu; Wang, Di; Chen, Xiaoyan; Wang, Huaxiang

doi:10.1109/jsen.2022.3161025

Cited by 16 publications

(7 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The simulated results of the media distribution reconstruction based on CV multi-frequency weighted method are shown in tables 4, 6, 8, 10, 12, 14 and the single frequency method is also applied for comparison. The reconstructed results are quantitatively evaluated by the image correlation coefficient [36], and shown in…”

Section: Results Analysismentioning

confidence: 99%

A CV weighted method for oil-water two-phase flow of non-contact multi-frequency el7ectrical impedance tomography

Bai,

Wang

2024

Phys. Scr.

View full text Add to dashboard Cite

Electrical tomography (ET) technology has attracted extensive attention in the field of scientific research on oil-water two-phase flow due to its advantages of non-intrusion, non-radiation, and simple equipment structure. Affected by Maxwell-Wagner-Sillars (MWS) effect, more imaging information can be obtained by changing the excitation frequency, especially for the mixtures with conductive phases. In this paper, a CV (coefficient of variation )-weighted method of non-contact multi-frequency electrical impedance tomography (NM-EIT) system is carried out for water-oil two-phase flow imaging. Firstly, the frequency response of the measurement system is studied and the electrical impedance information of medium in the region of interest (RoI) is observed. Secondly, the measured electrical impedance information with with different excitation frequencies are weighted by CV method. Finally, combined with the classic image reconstruction algorithm, the simulation and static experiment are carried out to verify the reconstruction effect of the proposed method. The imaging accuracies under different frequencies and flow patterns are analyzed, and a measurement experimental platform is set up to verify the reconstruction effect of the proposed method. The experimental results show that the CV-based multi-frequency weighted method maintains a higher imaging accuracy than single-frequency imaging.

show abstract

Section: Results Analysismentioning

confidence: 99%

A CV weighted method for oil-water two-phase flow of non-contact multi-frequency el7ectrical impedance tomography

Bai,

Wang

2024

Phys. Scr.

View full text Add to dashboard Cite

show abstract

“…This result indicated that the purpose of early research on EIT hardware was to improve the EIT image quality [‘contrast’ and ‘distinguishability’ ( Adler et al, 2011 ) ] through the cooperation of hardware (‘electrode model’ ( Mueller et al, 1999 )) and algorithm (‘state estimation’ ( Kolehmainen et al, 2001 ) and ‘decomposition’ ( Clay and Ferree, 2002 )). Recent keywords mainly focus on the performance improvement of each module of EIT hardware system by applied new designs or techniques, including ‘voltage measurement’ ( Fu et al, 2022 ; Shi et al, 2022 ), current source (‘Howland current source’ ( Oh et al, 2011 ), ‘current driver’ ( Hong et al, 2010 ; Shahghasemi, 2020 )and ‘output impedance’ ( Shishvan et al, 2021 )) and novel electrode sensor (‘textile electrodes ( Katashev et al, 2018 ; Hu et al, 2021 )’ and ‘active electrode’ ( Wu et al, 2019 )). For example, Saulnier et al and Liu et al designed a DSP-based current source and a FPGA-based adaptive different current source in order to meet the requirement of high output impedance for current source in the most popular EIT hardware frameworks with a parallel and multiple source architecture ( Saulnier et al, 2020 ; Liu et al, 2021 ).…”

Section: Discussionmentioning

confidence: 99%

Characteristics and topic trends on electrical impedance tomography hardware publications

Qin

Yao²,

Xu³

et al. 2022

Front. Physiol.

View full text Add to dashboard Cite

Objective: Electrical impedance tomography (EIT) is a technique to measure electrical properties of tissue. With the progress of modern integrated circuits and microchips, EIT instrumentation becomes an active research area to improve all aspects of device performance. Plenty of studies on EIT hardware have been presented in prestigious journals. This study explores publications on EIT hardware to identify the developing hotspots and trends.Method: Publications covering EIT hardware on the Web of Science Core Collection (WoSCC) database from 1989 to 2021 were collected for bibliometric analysis. CiteSpace and VOS viewer were used to study the characteristics of the publications.Main results: A total of 592 publications were analyzed, showing that the number of annual publications steadily increased. China, England, and South Korea were the most prolific countries on EIT hardware publications with productive native institutions and authors. Research topics spread out in “bio-electrical impedance imaging”, “hardware optimization”, “algorithms” and “clinical applications” (e.g., tissue, lung, brain, and oncology). Hardware research in “pulmonary” and “hemodynamic” applications focused on monitoring and were represented by silhouette recognition and dynamic imaging while research in “tumor and tissue” and “brain” applications focused on diagnosis and were represented by optimization of precision. Electrode development was a research focus through the years. Imaging precision and bioavailability of hardware optimization may be the future trend.Conclusion: Overall, system performance, particularly in the areas of system bandwidth and precision in applications may be the future directions of hardware research.

show abstract

“…In addition, convolutional neural networks (CNN)s also play a role in improving the imaging quality in the data-driven imaging framework, such as Li et al proposed V-Net [39], Li et al added dense connectivity to a deep CNN and designed VD-Net [40] to implement electrical resistance tomography. Zheng et al used CNNs as feature encoders and then used fully connected distribution vectors [41] to reconstruct electrical capacitance tomography, Zheng et al proposed an improved VGG-RBF to reconstruct lung EIT images [41], Fu et al combined the L2 regularization with CNN feature encoding to design a parallel regularizationguided depth imaging model [42], and Li et al composed the Land-weber iterative framework as a deep Unrolled framework to achieve EIT image reconstruction [43]. The abovementioned deep imaging models do not require forward operators, but all of them depend on abundant training samples.…”

Section: Related Workmentioning

confidence: 99%

HybridDenseU-Net: learning a multi-scale convolution and dense connectivity CNN for inverse imaging problems

Zhang,

Wang,

Chen

et al. 2023

Meas. Sci. Technol.

Self Cite

View full text Add to dashboard Cite

Inverse imaging problems (IIPs) is a cutting-edge technology which is part of the nonlinear inverse problem, the solution approaches to which have placedattention on deep learning recently. This paper proposes a unique learning-based framework for IIPs, referred to as HybridDenseU-Net, which takes U-Net as the backbone and optimizes the encoder as a two-branch feature extraction module. Compared to the direct skip-connection in conventional U-Net, dense connections are introduced to merge features between feature maps with the same dimension and construct multi-scale content in the decoder. The validation of HybridDenseU-Net is carried out by a case study of electrical impedance tomography, which is of typical nonlinear IIP. The results illustrate that HybridDenseU-Net has root mean square error of 3.0867 and structural similarity of 0.9846, which are significantly better than some state-of-the-art deep learning-based frameworks. It has been proven that this work could provide a promising idea for future research on learning-based image reconstruction methods.

show abstract

A Regularization-Guided Deep Imaging Method for Electrical Impedance Tomography

Cited by 16 publications

References 37 publications

A CV weighted method for oil-water two-phase flow of non-contact multi-frequency el7ectrical impedance tomography

A CV weighted method for oil-water two-phase flow of non-contact multi-frequency el7ectrical impedance tomography

Characteristics and topic trends on electrical impedance tomography hardware publications

HybridDenseU-Net: learning a multi-scale convolution and dense connectivity CNN for inverse imaging problems

Contact Info

Product

Resources

About