A review of errors in multi-frequency EIT instrumentation

An Approach to Using Electrical Impedance Myography Signal Sensors to Assess Morphofunctional Changes in Tissue during Muscle Contraction

This present work is aimed at conducting fundamental and exploratory studies of the mechanisms of electrical impedance signal formation. This paper also considers morphofunctional changes in forearm tissues during the performance of basic hand actions. For this purpose, the existing research benches were modernized to conduct experiments of mapping forearm muscle activity by electrode systems on the basis of complexing the electrical impedance signals and electromyography signals and recording electrode systems' pressing force using force transducers. Studies were carried out with the involvement of healthy volunteers in the implementation of vertical movement of the electrode system and ultrasound transducer when the subject's upper limb was positioned in the bed of the stand while performing basic hand actions in order to identify the relationship between the morphofunctional activity of the upper limb muscles and the recorded parameters of the electro-impedance myography signal. On the basis of the results of the studies, including complex measurements of neuromuscular activity on healthy volunteers such as the signals of electro-impedance myography and pressing force, analyses of the morphofunctional changes in tissues during action performance on the basis of ultrasound and MRI studies and the factors influencing the recorded signals of electro-impedance myography are described. The results are of fundamental importance and will enable reproducible electro-impedance myography signals, which, in turn, allow improved anthropomorphic control.

ACT5 Electrical Impedance Tomography System

OBJECTIVE

This article introduces the Adaptive Current Tomograph 5 (ACT5) Electrical Impedance Tomography (EIT) system. ACT5 is a 32 electrode applied-current multiple-source EIT system that can display real-time images of conductivity and susceptivity at 27 frames per second. The adaptive current sources in ACT5 can apply fully programmable current patterns with frequencies varying from 5 kHz to 500 kHz. The system also displays real-time ECG readings during the EIT imaging process.

METHODS

The hardware and software design and specifications are presented, including the current source design, FPGA hardware, safety features, calibration, and shunt impedance measurement.

RESULTS

Images of conductivity and susceptivity are presented from ACT5 data collected on tank phantoms and a human subject illustrating the system's ability to provide real-time images of pulsatile perfusion and ECG traces.

SIGNIFICANCE

The portability, high signal-to-noise ratio, and flexibility of applied currents over a wide range of frequencies enable this instrument to be used to obtain useful human subject data with relative clinical ease.

Metal Particle Detection by Integration of a Generative Adversarial Network and Electrical Impedance Tomography (GAN-EIT) for a Wet-Type Gravity Vibration Separator

The minor copper (Cu) particles among major aluminum (Al) particles have been detected by means of an integration of a generative adversarial network and electrical impedance tomography (GAN-EIT) for a wet-type gravity vibration separator (WGS). This study solves the problem of blurred EIT reconstructed images by proposing a GAN-EIT integration system for Cu detection in WGS. GAN-EIT produces two types of images of various Cu positions among major Al particles, which are (1) the photo-based GAN-EIT images, where blurred EIT reconstructed images are enhanced by GAN based on a full set of photo images, and (2) the simulation-based GAN-EIT images. The proposed metal particle detection by GAN-EIT is applied in experiments under static conditions to investigate the performance of the metal detection method under single-layer conditions with the variation of the position of Cu particles. As a quantitative result, the images of detected Cu by GAN-EIT ψ̿GAN in different positions have higher accuracy as compared to σ*EIT. In the region of interest (ROI) covered by the developed linear sensor, GAN-EIT successfully reduces the Cu detection error of conventional EIT by 40% while maintaining a minimum signal-to-noise ratio (SNR) of 60 [dB]. In conclusion, GAN-EIT is capable of improving the detailed features of the reconstructed images to visualize the detected Cu effectively.

Compensation for Electrode Detachment in Electrical Impedance Tomography with Wearable Textile Electrodes

Electrical impedance tomography (EIT) is a radiation-free and noninvasive medical image reconstruction technique in which a current is injected and the reflected voltage is received through electrodes. EIT electrodes require good connection with the skin for data acquisition and image reconstruction. However, detached electrodes are a common occurrence and cause measurement errors in EIT clinical applications. To address these issues, in this study, we proposed a method for detecting faulty electrodes using the differential voltage value of the detached electrode in an EIT system. Additionally, we proposed the voltage-replace and voltage-shift methods to compensate for invalid data from the faulty electrodes. In this study, we present the simulation, experimental, and in vivo chest results of our proposed methods to verify and evaluate the feasibility of this approach.

Parasitic Effects on Electrical Bioimpedance Systems: Critical Review

Parasitic capacitance represents the main error source in measurement systems based on electrical impedance spectroscopy. The capacitive nature of electrodes' impedance in tetrapolar configuration can give origin to phase errors when electrodes are coupled to parasitic capacitances. Nevertheless, reactive charges in tissue excitation systems are susceptible to instability. Based on such a scenario, mitigating capacitive effects associated with the electrode is a requirement in order to reduce errors in the measurement system. A literature review about the main compensation techniques for parasitic capacitance was carried out. The selected studies were categorized into three groups: (i) compensation in electronic instrumentation; (ii) compensation in measurement processing, and (iii) compensation by negative impedance converters. The three analyzed methods emerged as effective against fixed capacitance. No method seemed capable of mitigating the effects of electrodes' capacitance, that changes in the frequency spectrum. The analysis has revealed the need for a method to compensate varying capacitances, since electrodes' impedance is unknown.

Dynamic Hand Gesture Recognition Using Electrical Impedance Tomography

Electrical impedance tomography (EIT) has been applied in the field of human-computer interaction due to its advantages including the fact that it is non-invasive and has both low power consumption and a low cost. Previous work has focused on static gesture recognition based on EIT. Compared with static gestures, dynamic gestures are more informative and can achieve more functions in human-machine collaboration. In order to verify the feasibility of dynamic gesture recognition based on EIT, a traditional excitation drive pattern is optimized in this paper. The drive pattern of the fixed excitation electrode is tested for the first time to simplify the measurement process of the dynamic gesture. To improve the recognition accuracy of the dynamic gestures, a dual-channel feature extraction network combining a convolutional neural network (CNN) and gated recurrent unit (GRU), namely CG-SVM, is proposed. The new center distance loss is designed in order to simultaneously supervise the intra-class distance and inter-class distance. As a result, the discriminability of the confusing data is improved. With the new excitation drive pattern and classification network, the recognition accuracy of different interference data has increased by 2.7~14.2%. The new method has stronger robustness, and realizes the dynamic gesture recognition based on EIT for the first time.

Signal-to-noise ratio variance impact on the image reconstruction of electrical resistance tomography in solutions with high background conductivity

Electrical Resistance Tomography (ERT) has the potentialities of non-intrusive techniques and high temporal resolution which are essential characteristics for multiphase flow measurements. However, high background conductivities, such as saline water in oil extraction, impose a limitation in ERT image reconstruction. Focusing on the operational limits of an ERT tomography system operating in different conductivity backgrounds from 0.010 to 4.584 S/m, the impact on the image reconstruction was assessed via signal-to-noise variance. The signal-to-noise ratio (SNR) variance had a strong correlation (p-value = 5.40 × 10^-15) with the image reconstruction quality at the threshold of 30 dB, reaching a correlation value of r = -0.92 in the range of 0.010-0.246 S/m. Regarding the position error of the phantom, p-value = 1.30 × 10^-5 and r = -0.66 were attained. The global results revealed that the correlation of the mean of the SNR (p-value = 5 × 10^-4 and r = 0.55) was kept unaltered through the whole conductivity range, showing that such a statistical index can induce bias in establishing the operational limits of the hardware.

Detection of subclinical hemorrhage using electrical impedance: a porcine study

OBJECTIVE

Analyze the performance of electrical impedance tomography (EIT) in an innovative porcine model of subclinical hemorrhage and investigate associations between EIT and hemodynamic trends.

APPROACH

Twenty-five swine were bled at slow rates to create an extended period of subclinical hemorrhage during which the animal's heart rate (HR) and blood pressure (BP) remained stable from before hemodynamic deterioration, where stable was defined as < 15% decrease in BP and < 20% increase in HR - i.e. hemorrhages were hidden from standard vital signs of HR and BP. Continuous vital signs, photo-plethysmography, and continuous non-invasive EIT data were recorded and analyzed with the objective of developing an improved means of detecting subclinical hemorrhage - ideally as early as possible.

MAIN RESULTS

Best area-under-the-curve (AUC) values from comparing bleed to no-bleed epochs were 0.96 at a 80 ml bleed (~15.4 minutes) using an EIT-data-based metric and 0.79 at a 120 ml bleed (~23.1 minutes) from invasively measured BP - i.e. the EIT-data-based metric achieved higher AUCs at earlier points compared to standard clinical metrics without requiring image reconstructions.

SIGNIFICANCE

In this clinically relevant porcine model of subclinical hemorrhage, EIT appears to be superior to standard clinical metrics in early detection of hemorrhage.

A Wearable Bioimpedance Chest Patch for Real-Time Ambulatory Respiratory Monitoring

OBJECTIVE

This paper aims to introduce a wearable solution and a low-complexity algorithm for real-time continuous ambulatory respiratory monitoring.

METHODS

A wearable chest-worn patch is designed using a bioimpedance (BioZ) sensor to measure the changes in chest impedance caused by breathing. Besides, a medical-grade infrared temperature sensor is utilized to monitor body temperature. The computing algorithm implemented on the patch enables computation of breath-by-breath respiratory rate and chest temperature in real-time. Two wireless communication protocols are included in the system, namely Bluetooth and Long Range (LoRa), which enable both short-range and long-range data transmission.

RESULTS

The breathing rate measured in static (i.e., standing, sitting, supine, and lateral lying) and dynamic (i.e., walking, running, and cycling) positions by our device yielded an accuracy of more than 97.8% and 98.5% to the ground truth, respectively. Additionally, the devices performance is evaluated in real-world scenarios both indoors and outdoors.

CONCLUSION

The proposed system is capable of measuring breathing rate throughout a variety of daily activities. To the best of our knowledge, this is the first BioZ-based wearable patch capable of detecting breath-by-breath respiratory rate in real-time remotely under unrestricted ambulatory conditions.

SIGNIFICANCE

This study establishes a strategy for continuous respiratory monitoring that could aid in the early detection of cardiopulmonary disorders in everyday life.

A Multi-Frequency Focused Impedance Measurement System Based on Analogue Synchronous Peak Detection

Monitoring of anatomical structures and physiological processes by electrical impedance has attracted scientists as it is noninvasive, nonionizing and the instrumentation is relatively simple. Focused Impedance Method (FIM) is attractive in this context, as it has enhanced sensitivity at the central region directly beneath the electrode configuration minimizing contribution from neighboring regions. FIM essentially adds or averages two concentric and orthogonal combinations of conventional Tetrapolar Impedance Measurements (TPIM) and has three versions with 4, 6, and 8 electrodes. This paper describes the design and testing of a multi-frequency FIM (MFFIM) system capable of measuring all three versions of FIM at 8 frequencies in the range 10 kHz—1 MHz. A microcontroller based multi-frequency signal generator and a balanced Howland current source with high output impedance (476 kΩ at 10 kHz and 58.3 kΩ at 1 MHz) were implemented for driving currents into biological tissues with an error &lt;1%. The measurements were carried out at each frequency sequentially. The peak values of the amplified voltage signals were measured using a novel analogue synchronous peak detection technique from which the transfer impedances were obtained. The developed system was tested using TPIM measurements on a passive RC Cole network placed between two RC networks, the latter representing skin-electrode contact impedances. Overall accuracy of the measurement was very good (error &lt;4% at all frequencies except 1 MHz, with error 6%) and the resolution was 0.1 Ω. The designed MFFIM system had a sampling rate of &gt;45 frames per second which was deemed adequate for noninvasive real-time impedance measurements on biological tissues.

Electrical Impedance of Upper Limb Enables Robust Wearable Identity Recognition against Variation in Finger Placement and Environmental Factors

Most biometric authentication technologies commercialized in various fields mainly rely on acquired images of structural information, such as fingerprints, irises, and faces. However, bio-recognition techniques using these existing physical features are always at risk of template forgery threats, such as fake fingerprints. Due to the risk of theft and duplication, studies have recently been attempted using the internal structure and biological characteristics of the human body, including our previous works on the ratiometric biological impedance feature. However, one may still question its accuracy in real-life use due to the artifacts from sensing position variability and electrode-skin interfacing noise. Moreover, since the finger possesses more severe thermoregulatory vasomotion and large variability in the tissue properties than the core of the body, it is necessary to mitigate the harsh changes occurring at the peripheral extremities of the human body. To address these challenges, we propose a biometric authentication method through robust feature extraction from the upper-limb impedance acquired based on a portable wearable device. In this work, we show that the upper limb impedance features obtained from wearable devices are robust against undesirable factors such as finger placement deviations and day-to-day physiological changes, along with ratiometric impedance features. Overall, our upper-limb impedance-based analysis in a dataset of 1627 measurement from 33 subjects lowered the classification error rate from 22.38% to 4.3% (by a factor of 5), and further down to 2.4% (by a factor of 9) when combined with the ratiometric features.

Feasibility of Temperature Control by Electrical Impedance Tomography in Hyperthermia

We present a simulation study investigating the feasibility of electrical impedance tomography (EIT) as a low cost, noninvasive technique for hyperthermia (HT) treatment monitoring and adaptation. Temperature rise in tissues leads to perfusion and tissue conductivity changes that can be reconstructed in 3D by EIT to noninvasively map temperature and perfusion. In this study, we developed reconstruction methods and investigated the achievable accuracy of EIT by simulating HT treatmentlike scenarios, using detailed anatomical models with heterogeneous conductivity distributions. The impact of the size and location of the heated region, the voltage measurement signal-to-noise ratio, and the reference model personalization and accuracy were studied. Results showed that by introducing an iterative reconstruction approach, combined with adaptive prior regions and tissue-dependent penalties, planning-based reference models, measurement-based reweighting, and physics-based constraints, it is possible to map conductivity-changes throughout the heated domain, with an accuracy of around 5% and cm-scale spatial resolution. An initial exploration of the use of multifrequency EIT to separate temperature and perfusion effects yielded promising results, indicating that temperature reconstruction accuracy can be in the order of 1 ∘C. Our results suggest that EIT can provide valuable real-time HT monitoring capabilities. Experimental confirmation in real-world conditions is the next step.

An Imaged Based Method for Universal Performance Evaluation of Electrical Impedance Tomography Systems

This paper describes a simple and reproducible method for universal evaluation of the performance of electrical impedance tomography (EIT) systems using reconstructed images. To address the issues where common electrical parameters are not directly related to the quality of EIT images, based on objective full reference (FR) image quality assessment, the method provides a visually distinguishable hot colormap and two new FR metrics, the global and the more specific 'region of interest'. A passive 16 electrode EIT system using an application specific integrated circuit front-end was used to evaluate the proposed method. The measured results show, both visually and in terms of the proposed FR metrics, the impact on recorded EIT images with different design parameters and non-idealities. The paper also compares the image results of a passive electrode system with a matched 'single variable' active electrode system and demonstrates the merit of an active electrode system for noise interference. A figure of merit based on the FR metrics is proposed.

Wireless, Artefact Aware Impedance Sensor Node for Continuous Bio-Impedance Monitoring

Body bio-impedance is a unique parameter to monitor changes in body composition non-invasively. Continuous measurement of bio-impedance can track changes in body fluid content and cell mass and has widespread applications for physiological monitoring. State-of-the-art implementation of bio-impedance sensor devices is still limited for continuous use, in part, due to artefacts arising at the skin-electrode (SE) interface. Artefacts at the SE interface may arise due to various factors such as motion, applied pressure on the electrode surface, changes in ambient conditions or gradual drying of electrodes. This paper presents a novel bio-impedance sensor node that includes an artefact aware method for bio-impedance measurement. The sensor node enables autonomous and continuous measurement of bio-impedance and SE contact impedance at ten frequencies between 10 kHz to 100 kHz to detect artefacts at the SE interface. Experimental evaluation with SE contact impedance models using passive 2R1C electronic circuits and also with non-invasive in vivo measurements of SE contact impedance demonstrated high accuracy (with maximum error less than 1.5%) and precision of 0.6 Ω. The ability to detect artefacts caused by motion, vertically applied pressure and skin temperature changes was analysed in proof of concept experiments. Low power sensor node design achieved with 50mW in active mode and only 143 μW in sleep mode estimated a battery life of 90 days with a 250 mAh battery and duty-cycling impedance measurements every 60 seconds. Our method for artefact aware bio-impedance sensing is a step towards autonomous and unobtrusive continuous bio-impedance measurement for health monitoring at-home or in clinical environments.

A Review on Electrical Impedance Tomography Spectroscopy

Electrical Impedance Tomography Spectroscopy (EITS) enables the reconstruction of material distributions inside an object based on the frequency-dependent characteristics of different substances. In this paper, we present a review of EITS focusing on physical principles of the technology, sensor geometries, existing measurement systems, reconstruction algorithms, and image representation methods. In addition, a novel imaging method is proposed which could fill some of the gaps found in the literature. As an example of an application, EITS of ice and water mixtures is used.

DSP-based current source for electrical impedance tomography

OBJECTIVE

EIT systems, particularly those that use a parallel, multiple source architecture, require current sources with very high output impedance. To meet this requirement, sources often use complex analog circuits and require manual or electronically-controlled adjustments. The goal is to implement a current source with simple, adjustment-free analog electronics with high effective output impedance even with significant stray impedance at its output.

APPROACH

The excitation provided to the voltage-to-current converter is adjusted to accommodate the current lost in the finite output and stray impedances. The adaptive algorithm uses the measured voltage and the previously-measured output and stray impedance to determine the needed current adjustment.

MAIN RESULTS

The structure of the source is presented along with an implementation, and experimental results that show the effectiveness of the approach for frequencies up to 1 MHz. The measured output impedance with and without the adaptive compensation are presented as well as measurements of resistive and complex loads.

SIGNIFICANCE

The new current source has low analog complexity, operates over a wide range of frequencies, and can compensate for a significant stray shunt impedance. It can be used to implement improved parallel or serial EIT systems.

An Optimal Electrical Impedance Tomography Drive Pattern for Human-Computer Interaction Applications

In this article, we presented an optimal Electrical Impedance Tomography (EIT) drive pattern based on feature selection and model explanation, and proposed a portable EIT system for applications in human-computer interaction for gesture recognition and contact detection, which can reduce the measurement time and realize a performance trade-off between the accuracy and the time response. In our experiment, eleven hand gestures were designed to verify the proposed approach and EIT system. Compared to the traditional eight-electrode method, the optimal electrode drive pattern achieved a recognition accuracy of 97.5% with seven electrodes and the measurement time was reduced by 60%. To illustrate the universality of this method, we performed a contact detection experiment. By setting seven labels on the conductive panel and using optimal electrode drive pattern, the detection accuracy reached 100% with seven electrodes and the measurement time was reduced by 85%.

Smart Bioimpedance Spectroscopy Device for Body Composition Estimation

The purpose of this work is to describe a first approach to a smart bioimpedance spectroscopy device for its application to the estimation of body composition. The proposed device is capable of carrying out bioimpedance measurements in multiple configurable frequencies, processing the data to obtain the modulus and the bioimpedance phase in each of the frequencies, and transmitting the processed information wirelessly. Another novelty of this work is a new algorithm for the identification of Cole model parameters, which is the basis of body composition estimation through bioimpedance spectroscopy analysis. Against other proposals, the main advantages of the proposed method are its robustness against parasitic effects by employing an extended version of Cole model with phase delay and three dispersions, its simplicity and low computational load. The results obtained in a validation study with respiratory patients show the accuracy and feasibility of the proposed technology for bioimpedance measurements. The precision and validity of the algorithm was also proven in a validation study with peritoneal dialysis patients. The proposed method was the most accurate compared with other existing algorithms. Moreover, in those cases affected by parasitic effects the proposed algorithm provided better approximations to the bioimpedance values than a reference device.

A 122 fps, 1 MHz Bandwidth Multi-Frequency Wearable EIT Belt Featuring Novel Active Electrode Architecture for Neonatal Thorax Vital Sign Monitoring

A highly integrated, wearable electrical impedance tomography (EIT) belt for neonatal thorax vital multiple sign monitoring is presented. The belt has 16 active electrodes. Each electrode has an application-specific integrated circuit (ASIC) connected to it. The ASIC contains a fully differential current driver, a high-performance instrumentation amplifier, a digital controller, and multiplexors. The belt features a new active electrode architecture that allows programmable flexible electrode current drive and voltage sense patterns under simple digital control. It provides intimate connections to the electrodes for the current drive and to the IA for direct differential voltage measurement, providing superior common-mode rejection ratio. The ASIC was designed in a CMOS 0.35-μm high-voltage technology. The high-specification EIT belt has an image frame rate of 122 fps, a wide operating bandwidth of 1 MHz, and multi-frequency operation. It measures impedance with 98% accuracy and has less than 0.5 Ω and 1° variation across all possible channels. The image results confirmed the advantage of the new active electrode architecture and the benefit of wideband, multi-frequency EIT operation. The system successfully captured high-quality lung-respiration EIT images, breathing cycle, and heart rate. It can also provide boundary-shape information by using an array of MEMS sensors interfaced to the ASICs.

Current-conveyor-based wide-band current driver for electrical impedance tomography

OBJECTIVE

In this paper a wide-band integrated current driver for electrical impedance tomography (EIT) is presented. The application is primarily for prostate and breast cancer detection which require the tissue to be interrogated at frequencies up to 10 MHz while achieving low harmonic distortion and high accuracy.

APPROACH

The current driver is based on current conveyor architecture and can deliver 1.2 mA of peak to peak ac current between frequencies of 100 Hz-10 MHz. It is fabricated in CMOS 0.18 [Formula: see text]m technology with a power supply of 3.3 V, and occupies a core area of 0.26 [Formula: see text].

MAIN RESULTS

The measured harmonic distortion for a peak current of 1.2 mA is  <[Formula: see text] for frequencies less than 100 kHz, and increases to [Formula: see text] at 10 MHz. The measured output impedance of the current driver is 101 k[Formula: see text] at 1 MHz and 19.5 k[Formula: see text] at 10 MHz.

SIGNIFICANCE

The circuit is suitable for high frequency active electrode applications.

Structured Design Methodology to Achieve a High SNR Electrical Impedance Tomography

In this paper, we present a methodology for designing the main circuit building blocks of an electrical impedance tomography (EIT) system. In particular, we derive equations that map system-level EIT specifications to the performance requirements of each circuit block. We also review the circuit architectures that are best suited for meeting a given set of performance requirements. Our proposed design methodology is focused on maximizing the EIT system's signal-to-noise ratio while minimizing total power consumption.

bSSFP phase correction and its use in magnetic resonance electrical properties tomography

PURPOSE

Balanced steady-state free precession (bSSFP) sequence is widely used because of its high SNR and high speed. However, bSSFP images suffer from "banding artifact" caused by B₀ inhomogeneity. In this article, we propose a method to remove this artifact in bSSFP phase images and investigate the usage of the corrected phase images in phase-based magnetic resonance electrical properties tomography (MREPT).

THEORY AND METHODS

Two bSSFP phase images, obtained with different excitation frequencies, are collaged to get rid of the regions containing banding artifacts. Phase of the collaged bSSFP image is the sum of the transceive phase of the RF system and an error term that depends on B₀ and T₂ . By using B₀ and T₂ maps, this error is eliminated from bSSFP phase images by using pixel-wise corrections. Conductivity maps are obtained from the uncorrected and the corrected phase images using the phase-based cr-MREPT method.

RESULTS

Phantom and human experiment results of the proposed method are illustrated for both phase images and conductivity maps. It is shown that uncorrected phase images yield unacceptable conductivity images. When only B₀ information is used for phase correction conductivity, reconstructions are substantially improved, and yet T₂ information is still needed to fully recover accurate and undistorted conductivity images.

CONCLUSIONS

With the proposed technique, B₀ sensitivity of the bSSFP phase images can be removed by using B₀ and T₂ maps. It is also shown that corrected bSSFP phase images are of sufficient quality to be used in conductivity imaging.

Supervised Learning Classifiers for Electrical Impedance-based Bladder State Detection

Urinary Incontinence affects over 200 million people worldwide, severely impacting the quality of life of individuals. Bladder state detection technology has the potential to improve the lives of people with urinary incontinence by alerting the user before voiding occurs. To this end, the objective of this study is to investigate the feasibility of using supervised machine learning classifiers to determine the bladder state of ‘full’ or ‘not full’ from electrical impedance measurements. Electrical impedance data was obtained from computational models and a realistic experimental pelvic phantom. Multiple datasets with increasing complexity were formed for varying noise levels in simulation. 10-Fold testing was performed on each dataset to classify ‘full’ and ‘not full’ bladder states, including phantom measurement data. Support vector machines and k-Nearest-Neighbours classifiers were compared in terms of accuracy, sensitivity, and specificity. The minimum and maximum accuracies across all datasets were 73.16% and 100%, respectively. Factors that contributed most to misclassification were the noise level and bladder volumes near the threshold of ‘full’ or ‘not full’. This paper represents the first study to use machine learning for bladder state detection with electrical impedance measurements. The results show promise for impedance-based bladder state detection to support those living with urinary incontinence.

A novel post-processing scheme for two-dimensional electrical impedance tomography based on artificial neural networks

OBJECTIVE

Electrical Impedance Tomography (EIT) is a powerful non-invasive technique for imaging applications. The goal is to estimate the electrical properties of living tissues by measuring the potential at the boundary of the domain. Being safe with respect to patient health, non-invasive, and having no known hazards, EIT is an attractive and promising technology. However, it suffers from a particular technical difficulty, which consists of solving a nonlinear inverse problem in real time. Several nonlinear approaches have been proposed as a replacement for the linear solver, but in practice very few are capable of stable, high-quality, and real-time EIT imaging because of their very low robustness to errors and inaccurate modeling, or because they require considerable computational effort.

METHODS

In this paper, a post-processing technique based on an artificial neural network (ANN) is proposed to obtain a nonlinear solution to the inverse problem, starting from a linear solution. While common reconstruction methods based on ANNs estimate the solution directly from the measured data, the method proposed here enhances the solution obtained from a linear solver.

CONCLUSION

Applying a linear reconstruction algorithm before applying an ANN reduces the effects of noise and modeling errors. Hence, this approach significantly reduces the error associated with solving 2D inverse problems using machine-learning-based algorithms.

SIGNIFICANCE

This work presents radical enhancements in the stability of nonlinear methods for biomedical EIT applications.

The Frequency Spectral Properties of Electrode-Skin Contact Impedance on Human Head and Its Frequency-Dependent Effects on Frequency-Difference EIT in Stroke Detection from 10Hz to 1MHz

Frequency-difference electrical impedance tomography (fdEIT) reconstructs frequency-dependent changes of a complex impedance distribution. It has a potential application in acute stroke detection because there are significant differences in impedance spectra between stroke lesions and normal brain tissues. However, fdEIT suffers from the influences of electrode-skin contact impedance since contact impedance varies greatly with frequency. When using fdEIT to detect stroke, it is critical to know the degree of measurement errors or image artifacts caused by contact impedance. To our knowledge, no study has systematically investigated the frequency spectral properties of electrode-skin contact impedance on human head and its frequency-dependent effects on fdEIT used in stroke detection within a wide frequency band (10 Hz-1 MHz). In this study, we first measured and analyzed the frequency spectral properties of electrode-skin contact impedance on 47 human subjects’ heads within 10 Hz-1 MHz. Then, we quantified the frequency-dependent effects of contact impedance on fdEIT in stroke detection in terms of the current distribution beneath the electrodes and the contact impedance imbalance between two measuring electrodes. The results showed that the contact impedance at high frequencies (>100 kHz) significantly changed the current distribution beneath the electrode, leading to nonnegligible errors in boundary voltages and artifacts in reconstructed images. The contact impedance imbalance at low frequencies (<1 kHz) also caused significant measurement errors. We conclude that the contact impedance has critical frequency-dependent influences on fdEIT and further studies on reducing such influences are necessary to improve the application of fdEIT in stroke detection.

System Description and First Application of an FPGA-Based Simultaneous Multi-Frequency Electrical Impedance Tomography

A new prototype of a multi-frequency electrical impedance tomography system is presented. The system uses a field-programmable gate array as a main controller and is configured to measure at different frequencies simultaneously through a composite waveform. Both real and imaginary components of the data are computed for each frequency and sent to the personal computer over an ethernet connection, where both time-difference imaging and frequency-difference imaging are reconstructed and visualized. The system has been tested for both time-difference and frequency-difference imaging for diverse sets of frequency pairs in a resistive/capacitive test unit and in self-experiments. To our knowledge, this is the first work that shows preliminary frequency-difference images of in-vivo experiments. Results of time-difference imaging were compared with simulation results and shown that the new prototype performs well at all frequencies in the tested range of 60 kHz-960 kHz. For frequency-difference images, further development of algorithms and an improved normalization process is required to correctly reconstruct and interpreted the resulting images.

On the influence of spread constant in radial basis networks for electrical impedance tomography

Electrical impedance tomography (EIT) is a non-invasive imaging technique. The main task of this work is to solve a non-linear inverse problem, for which several techniques have been suggested, but none of which gives a very high degree of accuracy. This paper introduces a novel approach, based on radial basis function (RBF) artificial neural networks (ANNs), to solve this problem, and uses several ANNs to obtain the best solution to the EIT inverse problem. ANNs have the potential to directly estimate the solution of the inverse problem with a high degree of accuracy. While different radial basis neural networks do not always perform well on different problems, they usually give good results on some specific problems. This paper evidences a strong correlation between the area of the target and the spread constant of the RBF network that gives the best reconstruction. A solution to automatically estimate the size of the target and pick the best neural network directly from voltage measurements is presented, making the reconstruction process automatic. By automatically selecting the best ANN for each specific set of voltage measurements, the proposed solution gives a more accurate reconstruction of both small and large targets.

An impedance method for spatial sensing of 3D cell constructs--towards applications in tissue engineering

We present the characterisation and validation of multiplexed 4-terminal (4T) impedance measurements as a method for sensing the spatial location of cell aggregates within large three-dimensional (3D) gelatin scaffolds. The measurements were performed using an array of four rectangular chambers, each having eight platinum needle electrodes for parallel analysis. The electrode positions for current injection and voltage measurements were optimised by means of finite element simulations to maximise the sensitivity field distribution and spatial resolution. Eight different 4T combinations were experimentally tested in terms of the spatial sensitivity. The simulated sensitivity fields were validated using objects (phantoms) with different conductivity and size placed in different positions inside the chamber. This provided the detection limit (volume sensitivity) of 16.5%, i.e. the smallest detectable volume with respect to the size of the measurement chamber. Furthermore, the possibility for quick single frequency analysis was demonstrated by finding a common frequency of 250 kHz for all the presented electrode combinations. As final proof of concept, a high density of human hepatoblastoma (HepG2) cells were encapsulated in gelatin to form artificial 3D cell constructs and detected when placed in different positions inside large gelatin scaffolds. Taken together, these results open new perspectives for impedance-based sensing technologies for non-invasive monitoring in tissue engineering applications providing spatial information of constructs within biologically relevant 3D environments.

Multifrequency electrical impedance tomography with total variation regularization

Multifrequency electrical impedance tomography (MFEIT) reconstructs the distribution of conductivity by exploiting the dependence of tissue conductivity on frequency. MFEIT can be performed on a single instance of data, making it promising for applications such as stroke and cancer imaging, where it is not possible to obtain a 'baseline' measurement of healthy tissue. A nonlinear MFEIT algorithm able to reconstruct the volume fraction distribution of tissue rather than conductivities has been developed previously. For each volume, the fraction of a certain tissue should be either 1 or 0; this implies that the sharp changes of the fractions, representing the boundaries of tissue, contain all the relevant information. However, these boundaries are blurred by traditional regularization methods using [Formula: see text] norm. The total variation (TV) regularization can overcome this problem, but it is difficult to solve due to its non-differentiability. Because the fraction must be between 0 and 1, this imposes a constraint on the MFEIT method based on the fraction model. Therefore, a constrained optimization method capable of dealing with non-differentiable problems is required. Based on the primal and dual interior point method, we propose a new constrained TV regularized method to solve the fraction reconstruction problem. The noise performance of the new MFEIT method is analysed using simulations on a 2D cylindrical mesh. Convergence performance is also analysed through experiments using a cylindrical tank. Finally, simulations on an anatomically realistic head-shaped mesh are demonstrated. The proposed MFEIT method with TV regularization shows higher spatial resolution, particularly at the edges of the perturbation, and stronger noise robustness, and its image noise and shape error are 20% to 30% lower than the traditional fraction method.

A Reconstruction-Classification Method for Multifrequency Electrical Impedance Tomography

Multifrequency Electrical Impedance Tomography is an imaging technique which distinguishes biological tissues by their unique conductivity spectrum. Recent results suggest that the use of spectral constraints can significantly improve image quality. We present a combined reconstruction-classification method for estimating the spectra of individual tissues, whilst simultaneously reconstructing the conductivity. The advantage of this method is that a priori knowledge of the spectra is not required to be exact in that the constraints are updated at each step of the reconstruction. In this paper, we investigate the robustness of the proposed method to errors in the initial guess of the tissue spectra, and look at the effect of introducing spatial smoothing. We formalize and validate a frequency-difference variant of reconstruction-classification, and compare the use of absolute and frequency-difference data in the case of a phantom experiment.

A Sinusoidal Current Driver With an Extended Frequency Range and Multifrequency Operation for Bioimpedance Applications

This paper describes an alternative sinusoidal current driver suitable for bioimpedance applications where high frequency operation is required. The circuit is based on a transconductor and provides current outputs with low phase error for frequencies around its pole frequency. This extends the upper frequency operational limit of the current driver. Multifrequency currents can be generated where each individual frequency is phase corrected. Analysis of the circuit is presented together with simulation and experimental results which demonstrate the proof of concept for both single and dual frequency current drivers. Measurements on a discrete test version of the circuit demonstrate a phase reduction from 25° to 4° at 3 MHz for 2 mAp-p output current. The output impedance of the current driver is essentially constant at about 1.1 M Ω over a frequency range of 100 kHz to 5 MHz due to the introduction of the phase compensation. The compensation provides a bandwidth increase of a factor of about six for a residual phase delay of 4°.

Design of Bioimpedance Spectroscopy Instrument With Compensation Techniques for Soft Tissue Characterization

Bioimpedance spectroscopy (BIS) has shown significant potential in many areas of medicine to provide new physiologic markers. Several acute and chronic diseases are accompanied by changes in intra- and extracellular fluid within various areas of the human body. The estimation of fluid in various body compartments is therefore a simple and convenient method to monitor certain disease states. In this work, the design and evaluation of a BIS instrument are presented and three key areas of the development process investigated facilitating the BIS measurement of tissue hydration state. First, the benefit of incorporating DC-stabilizing circuitry to the standard modified Howland current pump (MHCP) is investigated to minimize the effect of DC offsets limiting the dynamic range of the system. Second, the influence of the distance between the bioimpedance probe and a high impedance material is investigated using finite element analysis (FEA). Third, an analytic compensation technique is presented to minimize the influence of parasitic capacitance. Finally, the overall experimental setup is evaluated through ex vivo BIS measurements of porcine spleen tissue and compared to published results. The DC-stabilizing circuit demonstrated its ability to maintain DC offsets at less than 650 μV through 100 kHz while maintaining an output impedance of 1 MΩ from 100 Hz to 100 kHz. The proximity of a bioimpedance probe to a high impedance material such as acrylic was shown to increase measured impedance readings by a factor of 4x as the ratio of the distance between the sensing electrodes to the distance between the bioimpedance probe and acrylic reached 1:3. The average parasitic capacitance for the circuit presented was found to be 712 ± 128 pF, and the analytic compensation method was shown to be able to minimize this effect on the BIS measurements. Measurements of porcine spleen tissue showed close correlation with experimental results reported in published articles. This research presents the successful design and evaluation of a BIS instrument. Specifically, robust measurements were obtained by implementing a DC-stabilized current source, investigating probe-material proximity issues and compensating for parasitic capacitance. These strategies were shown to provide tissue measurements comparable with published literature.

On the application of frequency selective common mode feedback for multifrequency EIT

Common mode voltages are frequently a problem in electrical impedance tomography (EIT) and other bioimpedance applications. To reduce their amplitude common mode feedback is employed. Formalised analyses of both current and voltage feedback is presented in this paper for current drives. Common mode effects due to imbalances caused by the current drives, the electrode connections to the body load and the introduction of the body impedance to ground are considered. Frequency selective narrowband common mode feedback previously proposed to provide feedback stability is examined. As a step towards multifrequency applications the use of narrowband feedback is experimentally demonstrated for two simultaneous current drives. Measured results using standard available components show a reduction of 62 dB for current feedback and 31 dB for voltage feedback. Frequencies ranged from 50 kHz to 1 MHz.

Detection of admittivity anomaly on high-contrast heterogeneous backgrounds using frequency difference EIT

This paper describes a multiple background subtraction method in frequency difference electrical impedance tomography (fdEIT) to detect an admittivity anomaly from a high-contrast background conductivity distribution. The proposed method expands the use of the conventional weighted frequency difference EIT method, which has been used limitedly to detect admittivity anomalies in a roughly homogeneous background. The proposed method can be viewed as multiple weighted difference imaging in fdEIT. Although the spatial resolutions of the output images by fdEIT are very low due to the inherent ill-posedness, numerical simulations and phantom experiments of the proposed method demonstrate its feasibility to detect anomalies. It has potential application in stroke detection in a head model, which is highly heterogeneous due to the skull.

Active electrode IC for EEG and electrical impedance tomography with continuous monitoring of contact impedance

The IC presented integrates the front-end for EEG and Electrical Impedance Tomography (EIT) acquisition on the electrode, together with electrode-skin contact impedance monitoring and EIT current generation, so as to improve signal quality and integration of the two techniques for brain imaging applications. The electrode size is less than 2 cm(2) and only 4 wires connect the electrode to the back-end. The readout circuit is based on a Differential Difference Amplifier and performs single-ended amplification and frequency division multiplexing of the three signals that are sent to the back-end on a single wire which also provides power supply. Since the system's CMRR is a function of each electrode's gain accuracy, an analysis is performed on how this is influenced by mismatches in passive and active components. The circuit is fabricated in 0.35 μm CMOS process and occupies 4 mm(2), the readout circuit consumes 360 μW, the input referred noise for bipolar EEG signal acquisition is 0.56 μVRMS between 0.5 and 100 Hz and almost halves if only EEG signal is acquired.

Exploratory study on the methodology of fast imaging of unilateral stroke lesions by electrical impedance asymmetry in human heads

Stroke has a high mortality and disability rate and should be rapidly diagnosed to improve prognosis. Diagnosing stroke is not a problem for hospitals with CT, MRI, and other imaging devices but is difficult for community hospitals without these devices. Based on the mechanism that the electrical impedance of the two hemispheres of a normal human head is basically symmetrical and a stroke can alter this symmetry, a fast electrical impedance imaging method called symmetrical electrical impedance tomography (SEIT) is proposed. In this technique, electrical impedance tomography (EIT) data measured from the undamaged craniocerebral hemisphere (CCH) is regarded as reference data for the remaining EIT data measured from the other CCH for difference imaging to identify the differences in resistivity distribution between the two CCHs. The results of SEIT imaging based on simulation data from the 2D human head finite element model and that from the physical phantom of human head verified this method in detection of unilateral stroke.

Direct observation of unstained biological specimens in water by the frequency transmission electric-field method using SEM

Scanning electron microscopy (SEM) is a powerful tool for the direct visualization of biological specimens at nanometre-scale resolution. However, images of unstained specimens in water using an atmospheric holder exhibit very poor contrast and heavy radiation damage. Here, we present a new form of microscopy, the frequency transmission electric-field (FTE) method using SEM, that offers low radiation damage and high-contrast observation of unstained biological samples in water. The wet biological specimens are enclosed in two silicon nitride (SiN) films. The metal-coated SiN film is irradiated using a focused modulation electron beam (EB) at a low-accelerating voltage. A measurement terminal under the sample holder detects the electric-field frequency signal, which contains structural information relating to the biological specimens. Our results in very little radiation damage to the sample, and the observation image is similar to the transmission image, depending on the sample volume. Our developed method can easily be utilized for the observation of various biological specimens in water.

Multifrequency electrical impedance tomography using spectral constraints

Multifrequency electrical impedance tomography (MFEIT) exploits the dependence of tissue impedance on frequency to recover an image of conductivity. MFEIT could provide emergency diagnosis of pathologies such as acute stroke, brain injury and breast cancer. We present a method for performing MFEIT using spectral constraints. Boundary voltage data is employed directly to reconstruct the volume fraction distribution of component tissues using a nonlinear method. Given that the reconstructed parameter is frequency independent, this approach allows for the simultaneous use of all multifrequency data, thus reducing the degrees of freedom of the reconstruction problem. Furthermore, this method allows for the use of frequency difference data in a nonlinear reconstruction algorithm. Results from empirical phantom measurements suggest that our fraction reconstruction method points to a new direction for the development of multifrequency EIT algorithms in the case that the spectral constraints are known, and may provide a unifying framework for static EIT imaging.

Modular architecture of a multi-frequency electrical impedance tomography system: design and implementation

Electrical impedance tomography (EIT) provides means of imaging the electrical properties distribution of biological tissues and fluids while impedance spectroscopy (IS) allows measuring their frequency response in a more global way. Both require precise and well-integrated instrumentation. In this work, we propose a modular architecture of a multi-frequency EIT (MfEIT) system which has capabilities in implementing both IS and MfEIT. First, IS performance is assessed in vivo using a cuff electrode implanted around the rodent cervical vagus nerve. Second, MfEIT performance is evaluated in vitro based on saline phantom experiments. Overall system allows addressing a wide range of applications and proves effective both in vitro and in vivo.

Simulation of a current source with a Cole-Cole load for multi-frequency electrical impedance tomography

An accurate current source is one of the keys in the hardware of Electrical impedance Tomography systems. Limitations appear mainly at higher frequencies and for non-simple resistive loads. In this paper, we simulate an improved Howland current source with a Cole-Cole load. Simulations comparing two different op-amps (THS4021 and OPA843) were performed at 1 kHz to 1 MHz. Results show that the THS4021 performed better than the OPA843. The current source with THS4021 reaches an output impedance of 20 MΩ at 1 kHz and above 320 kΩ at 1 MHz, it provides a constant and stable output current up to 4 mA, in the complete range of frequencies, and for Cole-Cole (resistive and capacitive) load.

An amplitude-to-time conversion technique suitable for multichannel data acquisition and bioimpedance imaging

In this paper we exploit the high timing resolution offered by microprocessors to develop an amplitude measurement approach that is convenient for high channel count portable sinusoidal recording systems such as the bioimpedance measurements used in impedance imaging. This approach reduces the number of components required per channel, reducing cost, size and power consumption compared to the traditional approaches. The setup uses two high performance comparators to convert amplitude difference to a timing difference. This is captured by a high speed microprocessor. A straightforward algorithm removes DC and timing offsets. We suggest three modes of operation: fast: less than one period of the input, normal: exactly one input period and high precision: multiple input periods. The mean signal-to-noise ratio was 40, 81, and 112.4 dB in fast, normal, and high precision mode respectively for a range of resistive loads.

Nanofiber web textile dry electrodes for long-term biopotential recording

Electrode properties are key to the quality of measured biopotential signals. Ubiquitous health care systems require long-term monitoring of biopotential signals from normal volunteers and patients in home or hospital environments. In these settings it is appropriate to use dry textile electrode networks for monitoring purposes, rather than the gel or saline-sponge skin interfaces used with Ag/AgCl electrodes. In this study, we report performance test results of two different electrospun conductive nanofiber webs, and three metal plated fabrics. We evaluated contact impedance, step response, noise and signal fidelity performance indices for all five dry electrodes, and compared them to those of conventional Ag/AgCl electrodes. Overall, we found nanofiber web electrodes matched Ag/AgCl electrode performance more closely than metal plated fabric electrodes, with the contact resistance and capacitance of Ag plated PVDF nanofiber web electrodes being most similar to Ag/AgCl over the 10 Hz to 500 kHz frequency range. We also observed that step responses of all three metal-plated fabrics were poorer than those for nanofiber web electrodes and Ag/AgCl. Further, noise standard deviation and noise power spectral densities were generally lower in nanofiber web electrodes than metal plated fabrics; and waveform fidelity of ECG-like traces recorded from nanofiber web electrodes was higher than for metal plated fabrics. We recommend textile nanofiber web electrodes in applications where flexibility, comfort and durability are required in addition to good electrical characteristics.

Biocompatible, high precision, wideband, improved Howland current source with lead-lag compensation

The Howland current pump is a popular bioelectrical circuit, useful for delivering precise electrical currents. In applications requiring high precision delivery of alternating current to biological loads, the output impedance of the Howland is a critical figure of merit that limits the precision of the delivered current when the load changes. We explain the minimum operational amplifier requirements to meet a target precision over a wide bandwidth. We also discuss effective compensation strategies for achieving stability without sacrificing high frequency output impedance. A current source suitable for Electrical Impedance Tomography (EIT) was simulated using a SPICE model, and built to verify stable operation. This current source design had stable output impedance of 3.3 MΩ up to 200 kHz, which provides 80 dB precision for our EIT application. We conclude by noting the difficulty in measuring the output impedance, and advise verifying the plausibility of measurements against theoretical limitations.

Modelling of an oesophageal electrode for cardiac function tomography

There is a need in critical care units for continuous cardiopulmonary monitoring techniques. ECG gated electrical impedance tomography is able to localize the impedance variations occurring during the cardiac cycle. This method is a safe, inexpensive and potentially fast technique for cardiac output imaging but the spatial resolution is presently low, particularly for central locations such as the heart. Many parameters including noise deteriorate the reconstruction result. One of the main obstacles in cardiac imaging at the heart location is the high impedance of lungs and muscles on the dorsal and posterior side of body. In this study we are investigating improvements of the measurement and initial conductivity estimation of the internal electrode by modelling an internal electrode inside the esophagus. We consider 16 electrodes connected around a cylindrical mesh. With the random noise level set near 0.05% of the signal we evaluated the Graz consensus reconstruction algorithm for electrical impedance tomography. The modelling and simulation results showed that the quality of the target in reconstructed images was improved by up to 5 times for amplitude response, position error, resolution, shape deformation and ringing effects with perturbations located in cardiac related positions when using an internal electrode.