电阻层析成像（ERT）技术详解

Introduction / 引言

In industrial inspection and medical diagnostics, there is a technology known as “electronic CT”—Electrical Resistance Tomography (ERT). It requires no X-rays, relies on no sound waves, and only by arranging electrodes on an object’s surface and applying weak currents can it “see through” to the internal conductivity distribution. With its unique advantages of non-invasiveness, no radiation, real-time imaging, and low cost, ERT has become one of the core members of Process Tomography. This article systematically reviews ERT technology from four aspects: basic principles, differences from ECT, industrial applications, and technical challenges.

I. Basic Principles of ERT / 一、ERT的基本原理

1.1 Conductivity Sensitivity Field / 电导率敏感场

The physical basis of ERT is very intuitive: different materials have different resistivities (the reciprocal of conductivity). For example, oil has extremely low conductivity (approximately 10⁻⁸~10⁻¹² S/m), while saltwater can have conductivity as high as tens of S/m. When we arrange a ring of electrodes around the measured object and inject alternating current from a pair of electrodes, the current will choose paths inside according to the conductivity distribution—higher current density in low-resistivity regions, lower in high-resistivity regions. This establishes a non-uniform voltage distribution within the sensitivity field.

The key is: if we can measure the voltages at the boundary electrode pairs and know the excitation conditions, then by solving the electromagnetic field inverse problem, we can reconstruct the internal conductivity distribution image. This is the core logic of ERT imaging.

1.2 Electrode Arrays and Data Acquisition Strategy / 电极阵列与数据采集策略

Taking the most common circular electrode array as an example: N electrodes (typically N=16 or 32) are arranged at equal intervals around the measured pipeline cross-section or human torso. The data acquisition process is as follows:

1. Excitation: Select a pair of adjacent or opposite electrodes to inject constant alternating current (typically sinusoidal current, frequency between 10kHz and 1MHz) to avoid electrode polarization effects.
2. Measurement: Successively measure boundary voltages between other electrode pairs to obtain a set of independent measurement values.
3. Switching: Change the position of the excitation electrode pair and repeat the excitation-measurement process until all independent excitation modes are covered.

For N electrodes using adjacent excitation mode, N(N-3)/2 independent voltage measurements can be obtained. The data acquisition system needs to complete the full scan in microseconds to milliseconds to meet real-time requirements.

1.3 Image Reconstruction / 图像重建

ERT image reconstruction is essentially a nonlinear inverse problem solving process. Commonly used methods include:

• Linear Back Projection (LBP): Fast computation but limited resolution, suitable for real-time qualitative monitoring.
• Sensitivity Matrix Method (Tikhonov regularization): Good balance between speed and accuracy, currently the mainstream in industrial applications.
• Newton-Raphson Iteration: High accuracy but large computational cost, suitable for offline fine analysis.
• Deep Learning Methods: In recent years, CNNs, U-Nets and other network architectures have been introduced, pre-trained with simulation data, showing excellent performance in flow pattern recognition tasks.

It should be noted that the ERT inverse problem is inherently ill-posed: small perturbations in measurement data can lead to dramatic changes in reconstructed images. Therefore, the choice of regularization strategy is crucial.

II. Differences and Connections Between ERT and ECT / 二、ERT与ECT的区别与联系

In the electrical tomography family, ERT is often discussed together with Electrical Capacitance Tomography (ECT). Both belong to “soft-field” imaging technology, sharing many methodological foundations, but there are significant differences in physical mechanisms and application scenarios.

Comparison Dimension	ERT	ECT
Sensitive Physical Quantity	Conductivity (σ)	Permittivity (ε)
Measurement Signal	Voltage (mV~V level)	Capacitance (fF~pF level)
Applicable Media	Primarily conductive media	Primarily non-conductive/insulating media
Typical Applications	Chemical multiphase flow (water-in-oil), geological exploration, EIT	Gas-solid fluidized beds, oil/gas two-phase flow, powder transport
Signal Noise Immunity	Relatively strong (large voltage signal)	Relatively weak (very small capacitance signal, susceptible to interference)
Electrode Material Requirements	Corrosion-resistant metals like stainless steel	Copper foil/copper rings, requiring shielding

In terms of connections: Their mathematical frameworks are highly similar—both based on the quasi-static approximation of Maxwell’s equations, both use circular electrode arrays, and image reconstruction algorithms can learn from each other. In certain complex multiphase flow scenarios, ERT and ECT can even form a dual-modality imaging system, simultaneously acquiring both conductivity and permittivity information for more comprehensive process monitoring.

III. Industrial and Medical Applications / 三、工业与医学应用

3.1 Chemical Process Monitoring / 化工过程监测

ERT is most maturely applied in the chemical industry. Taking a stirred reactor as an example, by installing an electrode array on the reactor wall, one can visualize internal mixing uniformity in real-time, monitor reactant concentration distribution, and identify short-circuit flows and dead zones. In multiphase flow pipelines, ERT can distinguish oil bubbles and gas bubble distributions in water-continuous phases, providing quantitative basis for flow pattern identification. Many domestic and international petrochemical enterprises have deployed online ERT monitoring systems in key equipment such as catalytic cracking units and oil-water separators.

3.2 Geological Exploration and Environmental Monitoring / 地质勘探与环境监测

In groundwater pollution monitoring, ERT is widely used to delineate the diffusion range and migration pathways of underground pollutants (such as heavy metals, oils). By arranging electrode arrays on the surface or in boreholes and conducting time-lapse ERT observations, three-dimensional evolution images of contaminant plumes can be obtained. In scenarios such as mine goaf detection and dam seepage detection, ERT is also an important non-destructive testing method.

3.3 Medical EIT: Electrical Impedance Tomography / 医学EIT：电阻抗断层成像

When ERT is applied to the human body, it is usually called Electrical Impedance Tomography (EIT). The most promising clinical applications of EIT include:

• Lung Ventilation Monitoring: By placing electrodes on the chest wall, real-time imaging of pulmonary ventilation distribution can be used for synchronized monitoring with ICU ventilators, avoiding the cumulative radiation risk of traditional X-ray examinations.
• Breast Cancer Screening: Using the difference in conductivity between cancerous and normal breast tissue for non-invasive detection.
• Brain Functional Imaging: Inferring neural activity regions by monitoring weak changes in brain conductivity, with the advantage of millisecond-level temporal resolution.

EIT has already obtained medical device certification and entered clinical use in some European countries.

3.4 Other Applications / 其他应用

The application landscape of ERT continues to expand: used in the food industry to monitor concentration processes; in ocean engineering for sediment detection; in civil engineering for concrete defect detection, and more.

IV. Key Technical Challenges and Development Trends / 四、关键技术挑战与发展趋势

Despite significant progress in ERT technology, several core challenges remain:

Limited spatial resolution. Constrained by electrode count and “soft-field” characteristics, ERT resolution is far inferior to X-ray CT, typically providing only coarse-grained regional distribution information. Increasing electrode count can improve resolution but increases circuit complexity and acquisition time.

Difficulty in 3D imaging. Most ERT systems currently focus on 2D cross-sectional imaging. 3D reconstruction requires significantly increasing electrode count and computational load, making real-time performance difficult to guarantee. Optimal arrangement of 3D electrode arrays is an active research direction.

Model-reality matching problems. Image reconstruction relies on the accuracy of the forward model. Factors such as irregular boundary shapes in actual applications and changes in electrode contact impedance can introduce model errors, affecting imaging quality.

Insufficient standardization. Lack of unified performance evaluation standards and calibration specifications makes results from different manufacturers and laboratories difficult to directly compare, somewhat constraining technology promotion.

In terms of development trends, the following directions deserve special attention:

First is deep learning and physical model fusion. Embedding neural networks as regularizers into traditional iterative reconstruction frameworks retains the robustness of physical constraints while gaining the high-resolution expressive power of data-driven approaches.

Second is multimodal fusion imaging. Combining ERT with ECT, ultrasound, fiber optic sensing and other technologies for complementary advantages has become an important development direction for high-end process imaging systems.

Third is introduction of flexible electronics. Using flexible printed circuit boards to create wearable electrode arrays greatly enhances human EIT comfort and signal consistency.

Fourth is edge computing and embedded systems. With increasing ARM and FPGA platform computing power, real-time 3D ERT imaging is moving from laboratory to industrial site.

Conclusion / 结语

From its conceptual proposal in the 1980s to today, ERT has gradually moved from pure academic research to industrial deployment and clinical application. It may never replace X-ray CT in resolution, but its unique value in safety, cost, real-time performance, and deployability makes it irreplaceable in scenarios such as process monitoring, environmental detection, and bedside monitoring. With continued progress in sensor technology, computing power, and artificial intelligence, ERT’s imaging capabilities and application breadth will continue to expand. For engineers and researchers focused on process control and advanced detection technology, ERT is undoubtedly a technology direction worth deep understanding and tracking.