A review on reconstruction algorithms and hardware implementation in electrical impedance tomography

The current medical imaging techniques can only be used in a few diagnostic scenarios after the development of qualitative lesions, and they frequently include the use of radiation, among other things. Electrical impedance tomography, in contrast, uses no radiation and is non-invasive. Electric impedance tomography (EIT), which has these benefits, is frequently utilized for the early stages of disease detection and treatment. This study discusses the research advancements, image reconstruction methods, hardware system design, and clinical applications of EIT in the treatment of lung disorders for the application of EIT in the treatment of lung lesions. The intricacy of EIT systems and their solutions is explained by looking at and introducing a few key components of EIT technology. This provides research ideas for future studies and confirms the technology’s extensive development prospects. The outcome demonstrates that EIT is still in a relatively early stage of development and that image reconstruction algorithms are now being utilized to improve imaging resolution. The accuracy of data collection and processing is increasing, and hardware technology is also advancing quickly. EIT is also employed in clinical settings for pathology in the bladder, brain, and lungs. Future uses of EIT in medicine have a lot of potential for real-time, long-term monitoring and early diagnosis.


Introduction
Electrical impedance tomography (EIT) is an imaging technique that uses electrical measurements to reconstruct images of the inside of objects [1].The principle of EIT is based on electrical impedance [2].Since Electrical Impedance Tomography can provide imaging of the interior of the human body in real time without harming the human body or putting it at risk, it is a promising method for noninvasive diagnostic medicine.EIT has mostly been employed in animal investigations and, hesitantly, in human clinical trials including digestive, respiratory, and cranial issues.Several researchers have employed the EIT approach in recent years to analyze soil characteristics, wood deterioration, root physiological condition, and plant cold tolerance.EIT imaging works by passing an electric current through the object being imaged and measuring the resulting voltage across the object.The conductivity and permittivity of the object can then be calculated from the voltage measurement.The internal structure of the item may be seen using these values.A material's composition and structure may be learned about by

Comparison of reconstruction algorithms
Electrical impedance tomography (EIT) is a promising medical imaging modality with a wide range of applications.The key to successful EIT imaging is having a robust and efficient reconstruction algorithm.There are many different reconstruction algorithms used in Electrical Impedance Tomography (EIT), each with its own strengths and weaknesses.The most common algorithm is the Filtered BackProjection (FBP) algorithm, which is fast and relatively simple to implement.However, FBP can be inaccurate in certain settings, like low-dose chest CT, leading to poor image quality.Filtered back-projection (FBP) reconstructions assume that every pixel represents attenuation accurately.FBP reconstructions compromise image quality due to the large variation in the value of these pixels caused by noise.Image noise or artefacts can be reduced by using Hybrid IR (HIR) [3].Based on the Fourier transform theory, the FBP algorithm is a space domain processing method.The reconstructed picture quality is improved, and the shape artefacts brought on by the point spread function are reduced as a result.It is characterized by convolution of the projection at each acquisition projection angle prior to inverse projection.
The Fourier central slicing theorem, which asserts that a one-dimensional Fourier transform of the projection is identical to a two-dimensional Fourier transform of the original picture, is the foundation of the procedure.A two-dimensional Fourier transform may be produced from each projection by applying the Fourier transform to it.The following steps can be taken to solve the problem of reconstructing the projection image: gather enough projections at various times (typically 180 acquisitions), determine the 1D Fourier transform of each projection, and combine the slices into a 2D Fourier transform of the image, and then use the Fourier inverse transform to obtain the reconstructed image.Newer algorithms have been developed that improve upon the accuracy of FBP while still being relatively fast and easy to implement.One such algorithm is the Sparse Reconstruction algorithm, which has been shown to produce more accurate images than FBP while still being computationally efficient [4].Other reconstruction algorithms include the Iterative Reconstruction algorithm, which can be more accurate than both FBP and Sparse Reconstruction but is also more computationally demanding.Finally, the Statistical Reconstruction algorithm is the most accurate of all the reconstruction algorithms but is also the slowest and most computationally intensive.Which reconstruction algorithm is best for a given EIT application depends on many factors, including the desired accuracy of the images, the computational resources available, and the time constraints of the application.
FBP is known to produce artifacts in the reconstructed image, especially near sharp boundaries [5].Another popular reconstruction algorithm is the Simultaneous Algebraic Reconstruction Technique (SART).SART is an iterative algorithm that can be used to reconstruct images from line integrals.SART is known to produce high-quality images, but it is also very slow [6].A third popular reconstruction algorithm is the Least Absolute Shrinkage and Selection Operator (LASSO).LASSO method regression can reduce multicollinearity and increase the accuracy of linear regression models.It is a convex optimization algorithm that can be used to reconstruct images from line integrals.LASSO is known to produce high-quality images, but it is also very slow [7].Finally, a fourth popular reconstruction algorithm is the Total Variation (TV) algorithm.TV is an optimization algorithm that can be used to reconstruct images from line integrals.TV is known to produce high-quality images, but it is also very slow.
Physiological data can be imaged in vivo using the electrical impedance tomography (EIT) image reconstruction algorithm, which has a regularization based on a total variation (TV) function.The reconstruction method aids in maintaining discontinuities in the reconstructed contours, such as stepwise variations in the electrical characteristics of inter-organ boundaries, which are not adequately addressed by more traditional optimization techniques, such as Newton's method.One of the newest reconstruction algorithms is the Sparse Reconstruction with Overlapping Constraints (SROC).SROC is a fast algorithm that can be used to reconstruct images from line integrals.SROC is known to produce highquality images with low artifacts.MBR is the newest and most promising reconstruction algorithm for EIT, but it is also the most computationally demanding.The MBR requires specialized hardware efficient implementation, but the results are worth it: high-quality reconstructions with minimal artifacts [8].As a result, the best reconstruction algorithm will be determined by the specific application and requirements.If a fast reconstruction with acceptable accuracy is required, FBP might be a good choice.If the highest possible accuracy is required, even at the expense of speed, then CO is your best option.If both are required, IR is the best compromise.It is important to understand the trade-off between speed and accuracy when choosing the algorithm that best suits your specific application.

Hardware implementation
The choice of which type of hardware to use for EIT reconstruction will depend on the specific requirements of the application: Dedicated reconstruction hardware is purpose-built for EIT and can offer high performance with low power consumption.However, these systems can be expensive and may not be compatible with all types of EIT systems.In terms of general-purpose processors, they are more widely available and often less expensive than dedicated reconstruction hardware.However, they may not offer the same level of performance or power efficiency [9].
There are many ways to implement Electrical Impedance Tomography (EIT), but one common factor among all implementations is the need for specialized hardware.This hardware can take many different forms, from dedicated EIT scanners to more general-purpose medical imaging devices that have been adapted for EIT.EIT can also be used with medical imaging tools like MRI and CT scanners.To meet these challenges, specialized EIT scanners were created.Arrays of electrodes placed on the body, as well as high-frequency signal generators and detectors, are commonly used [10].These scanners can be very expensive, however, so they are not always practical for clinical use.

Discussion
One challenge in EIT is that it requires very high-frequency electrical signals, on the order of millions of Hertz.This means that the hardware needs to be able to generate and detect these signals with great accuracy [11].Another challenge is that EIT images are often three-dimensional, so the hardware needs to be able to capture data from many electrodes placed around the body.The high-frequency signals are produced and detected in this instance using the already installed hardware, and the images are then created from the data using specialized software [12,13].Some commercial EIT systems use Finite Element Modelling (FEM) [11,14] for reconstruction, while others use BPA.There are also some research systems that use hybrid methods that combine both FEM and BPA.The choice of reconstruction algorithm will impact the design of the EIT system hardware.FEM-based systems will need more powerful processors and more memory, while BPA-based systems can be designed with simpler hardware.FEM and BMA have been implemented in hardware, but there are some important differences between the two.FEM requires more memory and processing power than BPA, but it can be more accurate.BPA is faster and easier to implement, but it can be less accurate [15].The magnitude, poor match, and time variability of the impedance of different imaging objects can lead to problems with resistive impedance tomography (EIT) measurements.The capacitance characteristics of the electrode interface, particularly those of human skin, are crucial at the high frequencies normally encountered in EIT.Currently, new methods can be employed to improve one component of the issue, including skin preparation, penetration enhancers, temperature, and electrical pulses.In the future, more comprehensive and refined techniques need to be explored in conjunction with cross-disciplines such as biomedical and agricultural sciences, and combining different techniques may be a viable way to minimize these problems.EIT technology is becoming increasingly mature in dynamic and static imaging, but there are still significant challenges in obtaining high quality static imaging compared to the currently well developed dynamic imaging.For example, it is limited by its ability to capture weak signals and the high threshold of reconstruction algorithms.Screening for diseases such as lung cancer and breast cancer precisely requires static imaging.
Electrical impedance imaging has the disadvantage of poor image quality, with insufficient spatial resolution and contrast.In order to meet the increasing medical needs in China, our medical equipment development capability is increasing, and electrical impedance imaging is receiving attention as a new medical imaging technology.Chinese scientists are committed to improving the image quality of electrical impedance imaging, and in 2018, a team from the Key Laboratory of Microscopic Magnetic Resonance of the Chinese Academy of Sciences developed a high-resolution, high-contrast dynamic electrical impedance image reconstruction algorithm, and in 2019, a high-resolution reconstruction of nondestructive medical electrical impedance images under several different imaging modalities was successfully achieved based on this algorithm.The efficient translation and marketability of the technical results have facilitated the development of the technology, and the pace of commercialization of electrical impedance imaging continues to accelerate as the technology continues to advance.Globally, the main manufacturers of electrical impedance imaging are Sentec (Switzerland), Swisstom (Switzerland), Maltron International (UK), Sciospec (Germany), Draeger (Germany), GE (USA) and other foreign companies, as well as domestic companies such as Hangzhou Yongchuan Technology Co Ltd and Silan Technology (Chengdu) Co Ltd.

Conclusion
Electrical impedance tomography is a powerful tool for imaging objects like the human body, plant roots, and aero-metals, and reconstruction algorithms and hardware implementation are critical for its success.In this review, we've looked at some of the most popular reconstruction algorithms and hardware platforms used in EIT.Overall, reconstruction algorithms and hardware implementation in Electrical Impedance Tomography play an important role in the accuracy of the imaging results.However, there is still room for improvement in terms of speed and efficiency.In addition, future research should focus on improving the quality of images produced by Electrical Impedance Tomography.With continued research and development, it is likely that these challenges will be addressed and that the potential of Electrical Impedance Tomography will be realized.Although the EIT system has been extensively studied in the last four decades, it remains a broad range of active research areas, with many other sectors to be applied.The application of EIT technology is still far from widespread.EIT has been employed in numerous medical applications, including lung ventilation monitoring, brain damage detection, bioimpedance analysis, and breast cancer screening.It is the only non-invasive bedside approach now available.But in applications such as crop health monitoring and non-destructive metal detection, there is a lack of innovation in theoretical research and implementation of intelligent monitoring systems.The review of this paper has some limitations, and more is to compare the advantages and disadvantages of different methods based on algorithmic principles and experimental prototypes based on physical principles.Future research in electrical tomography will primarily focus on acoustic tomography, medical imaging, multiphase flow dynamics in industrial processes, real-time monitoring of agricultural breeding and crop health, and novel diagnostic sensors for industrial and medical applications.and all my friends for their help and support.I could not have finished my thesis without all of their insightful advice and outstanding politeness.