Table of contents

This issue of Physiological Measurement follows the successful 13th ICEBI conference held at the Graz University of Technology, Austria, from 29 August to 2 September 2007. It was organized jointly with the 8th Conference on Electrical Impedance Tomography. The conference was co-organized by the Impedance Imaging Research Centre (IIRC) in Seoul and the Austrian Society for Biomedical Engineering (ÖGBMT), and it was kindly endorsed by the IFMBE. The combined conferences created a platform for investigators from both research communities of bio-impedance and EIT to engage in common areas of interest whilst also allowing an opportunity for the community to broaden its outlook in the areas of bio-sensors, clinical applications and new technologies. This upholds the tradition of successful conferences on biomedical applications of electrical impedance tomography and bio-impedance. It follows the 7th Conference on Biomedical Applications of Electrical Impedance Tomography combined with the World Congress 2006, which took place in Seoul from 27 August to 1 September 2006. The next EIT conference is scheduled to take place in Dartmouth College, USA, in June 2008.

This issue contains papers produced from discussion and feedback during the conference in both bio-impedance and EIT research areas. It was also an opportunity for new researchers to join the community and propose recent innovations. Of the 259 papers presented at the conference, Springer Verlag published 207 in the IFMBE proceedings. All authors were invited to prepare new papers for inclusion in this issue of Physiological Measurement. The manuscripts were put through a process of careful review before selection. A total of 43 were accepted, covering an important range of topics from bio-impedance, hardware, algorithms, new technologies and clinical applications.

From the scientific point of view, bio-impedance has a very long tradition that dates back to the days of Maxwell. Nevertheless, until the end of the 20th century, research was focused on the development of methods and basic experimental work while clinical or other practical applications remained limited. Consequently, there were not so many companies interested enough to produce professional equipment for easy and reliable data collection and interpretation. This may appear surprising as bio-impedance reflects so many (patho-) physiological processes, but on the other hand, a number of proposed applications, though sensitive, still exhibit low specificity, especially when aimed at processes far from the body surface. The 2007 conference may have shown a slight change of tendency. From 2000 to 2006, the number of papers cited in Medline and containing the keywords 'bio-impedance' or 'impedance tomography' increased by 56%. At the same time, we face an increasing number of applications related to micro- and nano-technologies that have emerged along with the tremendous growth of biochemical and cellular engineering. In recent years both the number of newly founded companies for bio-impedance devices and the involvement of established companies in bio-impedance research have increased.

The papers included in this year's issue clearly reflect this. New developments and trends are visible, such as non-contact methods using magnetic fields; MREIT, bringing together EIT and magnetic resonance imaging; and magnetic induction tomography (MIT), clinical applications, bio-impedance spectroscopy, new hardware and algorithms. The presentations of these new technologies continue to grow and it will be interesting to see how these contribute to future clinical applications.

At this conference, clinical applications were strongly represented; they included brain function, breast imaging, and thorax and gastric applications. It is important that researchers do not neglect the challenges of clinical applications of bio-impedance and EIT as there are still many technical difficulties that the technology needs to overcome in order to provide valuable clinical tools; however, there are promising signs that these tools are close to realization.

The future of both EIT and bio-impedance continues to provide researchers with new challenges. The high quality of research papers in this special issue shows clear evidence of significant advances in this research field.

In spontaneously breathing or ventilated subjects, it is difficult to image cardiac-related conductivity changes using electrical impedance tomography (EIT) due to the high amplitude of the ventilation component. Previous attempts to separate these components included either electrocardiogram-gated averaging, frequency domain filtering or holding the breath while performing the measurements. However, such methods are either not able to produce continuous real-time images or to fully separate cardiac and pulmonary changes. The aim of this work was to develop a new dynamic filtering method for the online separation of pulmonary and cardiac changes avoiding the drawbacks of the previous attempts. The approach is based on estimating template functions for the pulmonary and cardiac components by means of principal component analysis and frequency domain filtering. Then, these templates are fitted into the input signals. The new method enables an observer to examine the variation of the cardiac signal beat-by-beat after a one-time setup period of 20 s. Preliminary in vivo results of two healthy subjects are presented. The results are superior to frequency domain filtering and in good agreement with signals averaged over several cardiac cycles. The method does not depend on ECG or other a priori knowledge. The apparent validity of the method's ability to separate cardiac and pulmonary changes in EIT images was shown and has to be confirmed in future studies. The algorithm opens up new possibilities for future clinical trials on continuous monitoring by means of EIT and for the examination of the relation between the cardiac component and lung perfusion.

The paper presents a method for adaptive decomposition of an electrical bio-impedance (BI) signal into two components: cardiac and respiratory. The decomposition of a BI signal is not a trivial process because of the non-stationarity of the signal components and overlapping of their harmonic spectra. An application specific orthonormal basis (ASOB) was designed to solve the decomposition task using the Jacobi weighting function in the standard Gram–Schmidt process. The key element of the bio-impedance signal decomposer (BISD) is a model of the cardiac BI signal, which is constructed from the components of the ASOB and is intended for use in the BISD for on-line tracking of the cardiac BI signal. It makes it possible to separate the cardiac and respiratory components of the total BI signal in non-stationary conditions. In combination with the signal-shape locked loop (SSLL), the BISD allows us to decompose the BI signals with partially overlapping spectra. The proposed BISD based method is accomplished as a PC software digital system, but it is oriented towards applications in portable and stationary cardiac devices and in clinical settings.

Electrical impedance tomography (EIT) can be used to determine the admittivity distribution within the breast from measurements made on its surface. It has been reported that the electrical impedance spectrum of normal breast tissue is significantly different from that of malignant tissue, making EIT a candidate technology for breast cancer detection. The inhomogeneous structure of breasts, with thin low-admittivity skin layers covering the relatively high-admittivity tissue inside, makes the breast imaging problem difficult. In addition, studies show that the electrical properties of skin vary considerably over frequency. This paper proposes a layered forward model which incorporates the presence of skin. Our layered model has three layers, thin low-admittivity top and bottom layers representing skin and a thicker high-admittivity middle layer representing breast tissue. We solve for the forward solution of the layered geometry and compare its behavior with the previously used homogeneous model. Next we develop an iterative method to estimate the skin and breast tissue admittivities from the measured data, and study the robustness and accuracy of the method for various simulated and experimental data. We then look at the reconstruction of a target embedded in a layered body when the homogeneous forward solution is replaced by the layered forward solution. Lastly, we demonstrate the improvement that the layered forward model produces over the homogeneous model when working with clinical data.

Primarily this report outlines our investigation on utilizing magneto-acousto-electrical-tomography (MAET) to image the lead field current density in volume conductors. A lead field current density distribution is obtained when a current/voltage source is applied to a sample via a pair of electrodes. This is the first time a high-spatial-resolution image of current density is presented using MAET. We also compare an experimental image of current density in a sample with its corresponding numerical simulation. To image the lead field current density, rather than applying a current/voltage source directly to the sample, we place the sample in a static magnetic field and focus an ultrasonic pulse on the sample to simulate a point-like current dipole source at the focal point. Then by using electrodes we measure the voltage/current signal which, based on the reciprocity theorem, is proportional to a component of the lead field current density. In the theory section, we derive the equation relating the measured voltage to the lead field current density and the displacement velocity caused by ultrasound. The experimental data include the MAET signal and an image of the lead field current density for a thin sample. In addition, we discuss the potential improvements for MAET especially to overcome the limitation created by the observation that no signal was detected from the interior of a region having a uniform conductivity. As an auxiliary we offer a mathematical formula whereby the lead field current density may be utilized to reconstruct the distribution of the electrical impedance in a piecewise smooth object.

We present an improved approach to image ventilation in functional electrical impedance tomography (f-EIT). It combines the advantages of the two established procedures of calculating standard deviation as a functional parameter of ventilation (SD method) and the so-called filling capacity (FC method). The SD method quantifies the local impedance variation over a series of tomograms for each pixel; the FC method is based on the slope of a linear fit of regional versus the global impedance change. Tidal volume V_T is displayed linearly by the SD method in f-EIT; it is, however, sensitive to noisy data. The FC method is much more robust with respect to noise but does not display the tidal volume V_T. We combined the advantages of both techniques in a new VT method which is based on raw data. It saves computing time and is suitable for both f-EIT and absolute EIT (a-EIT). We separated the raw data into two representative sets: end expiratory and end inspiratory. This was accomplished by calculating the global time course of the relative impedance changes from the raw data. In this time course, we determined all frame numbers (indices) of end expiration and end inspiration. These frame numbers were used to calculate one mean expiratory and one mean inspiratory raw data frame. Reconstruction by difference imaging directly reflects the mean tidal volume V_T during the acquired frame series. The effect of the improvement by the VT method was investigated at different noise levels by adding artificial noise from 0 to 100 µV_rms to a real raw dataset. The robustness with regard to noise of the VT method was similar to that of the FC method. The practical value of suppression of non-ventilatory impedance changes, artefacts and noise was tested by studying ten healthy subjects (four females, six males) during normal breathing. We found a highly significant improvement in the image quality (p < 0.001) of ventilation for this group of volunteers.

Electrical impedance tomography (EIT) seeks to image the electrical conductivity of an object using electrical impedance measurement data at its periphery. Ultrasound reflection tomography (URT) is an imaging modality that is able to generate images of mechanical properties of the object in terms of acoustic impedance changes. Both URT and EIT have the potential to be used in various medical applications. In this paper we focus on breast tumour detection. Both URT and EIT belong to soft field tomography and suffer from the small amounts of available data and the inherently ill-posed nature of the inverse problems. These facts result in limited achievable reconstruction accuracy and resolution. A dual bio-electromechanical tomography system using ultrasound and electrical tomography is proposed in this paper to improve the detection of the small-size tumour. Data fusion techniques are implemented to combine the EIT/URT data. Based on simulations, we demonstrate the improvement of detection of small size anomalies and improved depth detection compared to single modality soft field tomography.

Electrical impedance tomography (EIT) reconstructs a conductivity change image within a body from electrical measurements on the body surface; while it has relatively low spatial resolution, it has a high temporal resolution. One key difficulty with EIT measurements is due to the movement and position uncertainty of the electrodes, especially due to breathing and posture change. In this paper, we develop an approach to reconstruct both the conductivity change image and the electrode movements from the temporal sequence of EIT measurements. Since both the conductivity change and electrode movement are slow with respect to the data frame rate, there are significant temporal correlations which we formulate as priors for the regularized image reconstruction model. Image reconstruction is posed in terms of a regularization matrix and a Jacobian matrix which are augmented for the conductivity change and electrode movement, and then further augmented to concatenate the d previous and future frames. Results are shown for simulation, phantom and human data, and show that the proposed algorithm yields improved resolution and noise performance in comparison to a conventional one-step reconstruction method.

Electrical impedance tomography (EIT) is very sensitive to deformations of the medium boundary shape. For lung imaging, breathing and changes in posture move the electrodes and change the chest shape, resulting in image artefacts. Several approaches have been proposed to improve the reconstructed images; most methods reconstruct both the boundary deformation and conductivity change from the measured data. These techniques require the calculation of the 'movement Jacobian', reflecting measurement changes due to the boundary deformation. Previous papers have calculated this Jacobian using perturbation techniques, which are slow (requiring multiple solutions of the forward problem) and become inaccurate with increasing finite element model size. This effect has limited reconstruction algorithms for deformable media to mostly 2D. To address this problem, we propose a direct method to calculate the Jacobian, based on a formulation of the derivatives of the finite element system matrix with respect to geometry changes. An illustrative example of these calculations is given, as well as a comparison between the proposed method and a perturbation method. Results show this method is ≈300 times faster; and for larger model sizes, the perturbation method begins to diverge from those from the direct method proposed.

We ask: how many bits of information (in the Shannon sense) do we get from a set of EIT measurements? Here, the term information in measurements (IM) is defined as: the decrease in uncertainty about the contents of a medium, due to a set of measurements. This decrease in uncertainty is quantified by the change from the inter-class model, q, defined by the prior information, to the intra-class model, p, given by the measured data (corrupted by noise). IM is measured by the expected relative entropy (Kullback–Leibler divergence) between distributions q and p, and corresponds to the channel capacity in an analogous communications system. Based on a Gaussian model of the measurement noise, Σ_n, and a prior model of the image element covariances Σ_x, we calculate IM , where [SNR]_i is the signal-to-noise ratio for each independent measurement calculated from the prior and noise models. For an example, we consider saline tank measurements from a 16 electrode EIT system, with a 2 cm radius non-conductive target, and calculate IM =179 bits. Temporal sequences of frames are considered, and formulae for IM as a function of temporal image element correlations are derived. We suggest that this measure may allow novel insights into questions such as distinguishability limits, optimal measurement schemes and data fusion.

We employed electrical impedance spectroscopy (EIS) to evaluate the electrical properties of prostatic tissues. We collected freshly excised prostates from 23 men immediately following radical prostatectomy. The prostates were sectioned into 3 mm slices and electrical property measurements of complex resistivity were recorded from each of the slices using an impedance probe over the frequency range of 100 Hz to 100 kHz. The area probed was marked so that following tissue fixation and slide preparation, histological assessment could be correlated directly with the recorded EIS spectra. Prostate cancer (CaP), benign prostatic hyperplasia (BPH), non-hyperplastic glandular tissue and stroma were the primary prostatic tissue types probed. Genetic and least squares parameter estimation algorithms were implemented for fitting a Cole-type resistivity model to the measured data. The four multi-frequency-based spectral parameters defining the recorded spectrum (ρ_∞, Δρ, f_c and α) were determined using these algorithms and statistically analyzed with respect to the tissue type. Both algorithms fit the measured data well, with the least squares algorithm having a better average goodness of fit (95.2 mΩ m versus 109.8 mΩ m) and a faster execution time (80.9 ms versus 13 637 ms) than the genetic algorithm. The mean parameters, from all tissue samples, estimated using the genetic algorithm ranged from 4.44 to 5.55 Ω m, 2.42 to 7.14 Ω m, 3.26 to 6.07 kHz and 0.565 to 0.654 for ρ_∞, Δρ, f_c and α, respectively. These same parameters estimated using the least squares algorithm ranged from 4.58 to 5.79 Ω m, 2.18 to 6.98 Ω m, 2.97 to 5.06 kHz and 0.621 to 0.742 for ρ_∞, Δρ, f_c and α, respectively. The ranges of these parameters were similar to those reported in the literature. Further, significant differences (p < 0.01) were observed between CaP and BPH for the spectral parameters Δρ and f_c; this is especially important since current prostate cancer screening methods do not reliably differentiate between these two tissue types.

Objective, non-invasive measures of lung maturity and development, oxygen requirements and lung function, suitable for use in small, unsedated infants, are urgently required to define the nature and severity of persisting lung disease, and to identify risk factors for developing chronic lung problems. Disorders of lung growth, maturation and control of breathing are among the most important problems faced by the neonatologists. At present, no system for continuous monitoring of neonate lung function to reduce the risk of chronic lung disease in infancy in intensive care units exists. We are in the process of developing a new integrated electrical impedance tomography (EIT) system based on wearable technology to integrate measures of the boundary diameter from the boundary form for neonates into the reconstruction algorithm. In principle, this approach could provide a reduction of image artefacts in the reconstructed image associated with incorrect boundary form assumptions. In this paper, we investigate the required accuracy of the boundary form that would be suitable to minimize artefacts in the reconstruction for neonate lung function. The number of data points needed to create the required boundary form is automatically determined using genetic algorithms. The approach presented in this paper is to assist quality of the reconstruction using different approximations to the ideal boundary form. We also investigate the use of a wavelet algebraic multi-grid (WAMG) preconditioner to reduce the reconstruction computation requirements. Results are presented that demonstrate a full 3D model is required to minimize artefact in the reconstructed image and the implementation of a WAMG for EIT.

Bioimpedance techniques may be appropriate for cardiac stroke volume (SV) monitoring since thoracic anatomical changes during the heart contraction reflect on the conductivity distribution. In some bioimpedance techniques, the electrical potential is calculated from the impedance distribution using Poisson's equation. That is called the forward problem and in many applications it is used inherently in the solution of the inverse problem—finding the impedance distribution from the electrical potentials. In this work, the forward problem was simulated using a realistic 3D hybrid phantom of the human thorax. The cardiac cycle of normal patients and patients suffering from cardiogenic pulmonary edema was simulated, including the effect of pulmonary blood perfusion during heart contraction. The forward problem was found to be most sensitive to SV when current was injected from the right breast toward the left scapula (−0.021 µV ml⁻¹). Our simulations show that both the heart volume and lung conductivity affect the developing voltage; therefore in SV estimation, the lung conductivity and heart volume should be jointly estimated.

The basic principles of electric field tomography (EFT) are briefly explained. EFT system numerical simulation results are given. The comparison of images reconstructed by systems with planar and round electrode arrays is presented. Some aspects of transmitting and measuring stages are discussed. The accuracy of the phase-sensitive demodulator, which is the key element of an EFT measuring system, is evaluated experimentally.

We present a simple method to determine systematic errors that will occur in the measurements by EIT systems. The approach is based on very simple scalable resistive phantoms for EIT systems using a 16 electrode adjacent drive pattern. The output voltage of the phantoms is constant for all combinations of current injection and voltage measurements and the trans-impedance of each phantom is determined by only one component. It can be chosen independently from the input and output impedance, which can be set in order to simulate measurements on the human thorax. Additional serial adapters allow investigation of the influence of the contact impedance at the electrodes on resulting errors. Since real errors depend on the dynamic properties of an EIT system, the following parameters are accessible: crosstalk, the absolute error of each driving/sensing channel and the signal to noise ratio in each channel. Measurements were performed on a Goe-MF II EIT system under four different simulated operational conditions. We found that systematic measurement errors always exceeded the error level of stochastic noise since the Goe-MF II system had been optimized for a sufficient signal to noise ratio but not for accuracy. In time difference imaging and functional EIT (f-EIT) systematic errors are reduced to a minimum by dividing the raw data by reference data. This is not the case in absolute EIT (a-EIT) where the resistivity of the examined object is determined on an absolute scale. We conclude that a reduction of systematic errors has to be one major goal in future system design.

Electrical impedance tomography (EIT) has many potential applications in medicine. The sensitivity and spatial resolution of EIT can be enhanced significantly by increasing the measurement signal-to-noise ratio (SNR), to which a significant contribution is made by the SNR of the current excitation sub-system. In this paper we present a prototype EIT current excitation sub-system with 80 dB SNR, exploiting both digital and analogue techniques. It uses digital waveform synthesis, a 16-bit DAC and subsequent reconstruction filter, to drive an enhanced Howland current source. Detailed analysis and testing of the current sub-system are presented. Its output impedance is 10 MΩ for different load impedances, varying slowly over the frequency range from 10 kHz up to 4 MHz.

Frequency domain impedance measurements are still the common approach in assessing passive electrical properties of cells and tissues. However, due to the time requirements for sweeping over a frequency range for performing spectroscopy, they are not suited for recovering fast impedance changes of biological objects. The use of broad bandwidth excitation and monitoring the response as a function of time will greatly reduce the measurement time. The widespread usage of a square wave excitation is simple but not always the best choice. Here we consider different waveforms for excitation and discuss not only the advantages but also their limitations. Measurements in a miniaturized chamber where frequency and time domain measurements are compared show the suitability of different waveforms as excitation signals for the measurements of bio-impedance. The chirp excitation has been found to be most promising in terms of frequency range, signal-to-noise ratio and crest factor.

The paper presents an autonomous programmable current generator module for multi-frequency EIT systems. The module incorporates all stages from the sine wave generation with frequency and amplitude tuning, D/A converter and filter, a high output resistance voltage-to-current converter to the associated digital communication and control. The paper presents in depth the original digital quadrature signal generator and the output current generator with a high resistance. The other main blocks of the design use current practice specifications, since recent technological solutions proposed in the literature were found appropriate. The proposed signal generator circuit, characterized by a very low complexity, is analyzed in its capacity to produce multiple accurate signals up to 1 MHz in frequency. The precision output current source uses a modified current conveyor of type CCII with a high output resistance and low distortion. The output current frequency spectrum and linearity parameters obtained in the simulations are also described. The simulation results indicate a good linearity and high output resistance with an acceptable output voltage swing. The calculated performance parameters are validated with simulations, and future work for the prototype fabrication of the IC is outlined.

A suspension of commercially available single-walled carbon nanotubes (SWNTs) is directly deposited onto a platinum multielectrode array surface. This is a novel and easy method to reduce interface impedance values which can be used instead of electromodified electrodes. This paper shows that this deposition method is a useful technique for the modification of patterned electrodes ranging in the micro scale. A thorough comparison between the common and well-known black platinum versus SWNTs, as interface material for different electrode areas, has been carried out. SWNTs-based electrodes smaller than 40 µm improve the interface impedance values when compared to black platinum-modified electrodes of the same size. The best results can be found for the 10 µm , which improves the electrode resistance by 25% in comparison with the black platinum ones. The lower resistance and higher capacitance calculated for the 40 µm diameter SWNTs-based electrode, in comparison with black platinum, also evidence a clear increment of the effective area, which is directly related to the impedance decrease.

Nanoscale probes have been developed for the online characterization of the electrical properties of biological cells by dielectric spectroscopy. Two types of sensors have been designed and fabricated. The first one is devoted to low (<10 MHz) frequency range analysis and consists of gold nanoelectrodes. The second one works for high (>40 Hz) frequency range analysis and consists of a gold nanowire. The patterning of the sensors is performed by electron beam lithography. These devices are integrated in a microfluidic channel network for the manipulation of the cells and for the improvement of the performances of the sensors. These devices are used for the analysis of a well-characterized biological model in the area of the ligand–receptor interaction. The purpose is to monitor the interaction between the lactoferrin (the ligand) and the nucleolin and sulfated proteoglycans (the receptors) present or not on a set of mutant Chinese hamster ovary cell lines and their following internalization into the cytoplasm. Initial measurements have been performed with this microsystem and they demonstrate its capability for label-free, real-time, analysis of a dynamic mechanism involving biological cells.

We report on the impedance mapping of in vitro cellular morphology by electrical impedance spectroscopy, using microelectrodes. A micro multielectrode system was designed, fabricated, assembled, tested and demonstrated for the monitoring of anchorage-dependent cell behavior and morphology. This system allowed continuous, label-free, quantitative monitoring and visualization of cell adhesion, spreading, proliferation and detachment due to cell cycle processes as well as cell–drug interaction, with spatio-temporal resolution. OvCa429 ovarian cancer cells were monitored in vitro over a period of 70 hours by inoculating the cell suspension directly on the multielectrode device. The phase angle of impedance was observed to develop a distinctive shape as a result of cell attachment and proliferation. The shape of the phase angle curve reverted back to the pre-attachment shape upon detachment of cells from the substrate, caused by the addition of trypsin to the cell culture medium. The impedance data of the cell culture were then successfully modeled as a multi-parametric equivalent circuit. The model incorporated both interfacial and cell-layer impedance parameters. Upon addition of trypsin, the cell-layer parameters showed a marked decline and were eventually eliminated from the multi-parametric model, confirming the correlation of the model to the electrode–cell–electrolyte system. These experiments demonstrate the applicability of the impedance mapping technique in visualizing and quantifying physiological changes in the cell layer due to cellular processes as well as the effect of external chemical stimulus on cells (cell–drug interaction).

The free beta subunit of human-chorionic-gonadotropin (hCGβ) is critical for various aspects of human health. Detection and quantification of this protein are essential during pregnancy as it provides clinicians valuable information regarding the progress of a pregnancy and the health of a foetus. Furthermore, it can be used as a biomarker for gestational trophoblastic disease (GTD), germ cell tumours and some non-trophoblastic gynaecological cancers and common epithelial tumours. Monitoring hCGβ levels is particularly important for patient treatment monitoring and relapse detection especially in GTD. This paper presents an investigation of the characteristics of the first two stages necessary for the development of a bio-impedance hCGβ sensor, using impedance spectroscopy and commercially available microelectrodes. Additionally, electrical equivalent circuit models based on the experimental results of these stages are presented. The biosensor is based on the formation of stable antibody–antigen complexes on golden microband electrodes covered with a layer of a self-assembled monolayer (SAM) or with both SAM and protein G. The preliminary results and analysis relate the interfacial processes and physical structure of the sensor to its electrical behaviour. Finally, preliminary results obtained from the sensor without protein G, which strongly indicate hCGβ detection, are also presented.

This numerical simulation study addressed the effects of the location of a discrete brain hematoma on the volumetric inductive phase shift of the brain measured with an induction circular sensor coil and an induction magnetron sensor coil. The theoretical study simulates the brain cavity as a circular sphere transversely centered with respect to the circular and magnetron sensor coils. As a case study for the effects of hematoma location, we employed similar size simulated spherical hematomas placed at three different positions from the center of the brain outward. A three-dimensional finite element analysis of the field equations in the frequency range from 100 kHz to 100 MHz revealed a substantial effect of hematoma location on the ability of both the circular and magnetron sensors to detect the hematomas. In particular it was found that there are frequencies, which may be related to resonance, at which the occurrence of the hematomas has no effect on the volumetric inductive phase shift of the brain. Furthermore it was found that the relative sensitivity of circular and magnetron sensor coils with respect to the occurrence of hematoma varies with the location of the hematoma.

The increasing number of applications of electrical bioimpedance measurements in biomedical practice, together with continuous advances in textile technology, has encouraged several researchers to make the first attempts to develop portable, even wearable, electrical bioimpedance measurement systems. The main target of these systems is personal and home monitoring. Analog Devices has made available AD5933, a new system-on-chip fully integrated electrical impedance spectrometer, which might allow the implementation of minimum-size instrumentation for electrical bioimpedance measurements. However, AD5933 as such is not suitable for most applications of electrical bioimpedance. In this work, we present a relatively simple analog front-end that adapts AD5933 to a four-electrode strategy, allowing its use in biomedical applications for the first time. The resulting impedance measurements exhibit a very good performance in aspects like load dynamic range and accuracy. This type of minimum-size, system-on-chip-based bioimpedance measurement system would lead researchers to develop and implement light and wearable electrical bioimpedance systems for home and personal health monitoring applications, a new and huge niche for medical technology development.

Bio-impedance measurements can be used to detect and monitor several properties of living hard-tissues, some of which include bone mineral density, bone fracture healing or dental caries detection. In this paper a simple method and hardware architecture for hard tissue bio-impedance measurement is proposed. The key design aspects of such architecture are discussed and a commercial handheld ac impedance device is presented that is fully certified to international medical standards. It includes a 4-channel multiplexer and is capable of measuring impedances from 10 kΩ to 10 MΩ across a frequency range of 100 Hz to 100 kHz with a maximum error of 5%. The device incorporates several user interface methods and a Bluetooth link for bi-directional wireless data transfer. Low-power design techniques have been implemented, ensuring the device exceeds 8 h of continuous use. Finally, bench test results using dummy cells consisting of parallel connected resistors and capacitors, from 10 kΩ to 10 MΩ and from 20 pF to 100 pF, are discussed.

Non-contact heart and lung activity monitoring would be a desirable supplement to conventional monitoring techniques. Based on the potential of non-contact magnetic induction measurements, requirements for an adequate monitoring system were estimated. This formed the basis for the development of the presented extendable multichannel simultaneous magnetic induction measurement system (MUSIMITOS). Special focus was given to the dynamic behaviour and simultaneous multichannel measurements, so that the system allows for up to 14 receiver coils working simultaneously at 6 excitation frequencies. Moreover, a real-time software concept for online signal processing visualization in combination with a fast software demodulation is presented. Finally, first steps towards a clinical application are pointed out and technical performance as well as first in vivo measurements are presented. This paper covers some aspects previously presented in Steffen and Leonhardt (2007 Proc. 13th Int. Conf. on Electrical Bioimpedance and the 8th Conf. on Electrical Impedance Tomography, Graz 2007).

Electrical impedance technology was used to characterize DNA recognition in a monolayer containing single-stranded DNA probes immobilized on a gold substrate using thiol self-assembly chemistry. Recognition of targeted complementary DNA was principally correlated with an eight-fold increase in the conductance of the monolayer and attributed to electron conduction through double helices formed upon the binding of the DNA targets to the probes. The high recognitive sensitivity was possible without the use of the redox labels or large bias voltages required for recognition using cyclic and Osteryoung square wave voltammetry. The impedance technology also provided atomic resolution of a hybrid bimolecular lipid membrane formed by deposition of a phospholipid:cholesterol monolayer onto a hydrophobic alkyl monolayer covalently attached to a silicon substrate via silicon–carbon bonds. Atomic resolution was achieved through preparation of membranes on surfaces approaching atomic flatness and the performance of impedance measurements over precisely defined areas of the surface in contact with solutions. Principally capacitive properties distinguished between the immobilized (octadecyl) and more fluidic (lipid:cholesterol) leaflets of the hybrid membrane. The lipid:cholesterol leaflets were structurally similar to those leaflets in free-standing bimolecular lipid membranes. The hybrid membrane therefore provides a highly stable and physiologically relevant surface for studying biomolecular interactions with membrane surfaces.

Dielectric properties of breast tissue obtained from 131 patients were investigated by an open-ended probe separated from the tissue by a dielectric film (plastic foil). The film decreases the measured capacitance, but the content of blood and body liquids has little effect on the measured quantity. We found that in cancerous tissue the electric capacity (related to the permittivity) measured at f ⩾ 10 MHz using the modified probe is higher than that in normal tissue. The dielectric assignment was confirmed by histopathology in 94% of investigated cases. High efficiency in differentiation between normal and cancer breast tissues and the feasibility of fulfilling of aseptic conditions (a film covering the probe) are promising in intra-operative cancer diagnostics.

Electrical impedance tomography (EIT) enables one to determine and visualize non-invasively the spatial distribution of the electrical properties of the tissues inside the body, thus providing valuable diagnostic information. The electrical impedance mammography (EIM) system is a specialized EIT system for diagnostics and imaging of the breast. While breast cancer is the main target for any investigation conducted in this area, the diagnosis of non-cancerous diseases is also very important because it opens the way to improve the quality of life for many women and it may also reduce the incidence of breast cancer through effective treatment of mastopathy. This paper presents the main results of a comprehensive examination of 166 women using four methods: multifrequency electrical impedance mammography, ultrasonic investigation, x-ray mammography and puncture biopsy. The objective of the investigation is to estimate the usefulness of multifrequency electrical impedance mammography for diagnosing dyshormonal mammary gland diseases. The results demonstrate the advantages of the multifrequency EIM method. In particular, dual-frequency electrical impedance mammography in contrast with the single-frequency variant enables one not only to diagnose mastopathy, but also allows accurate detection of its cystless form based on observation of the absence of any difference between average conductivity in both phases of the menstrual cycle. Because the cystless form of mastopathy is associated with a higher risk of cancer development, this method allows identification of a higher risk group of patients for more frequent investigations.

Extension of the frequency range of electrical impedance myography (EIM) to 2 MHz discloses a major rise in the reactance of muscle above 3–500 kHz, together with a slow decrease in the resistance consonant with the Kramers–Kronig relations. This 'upturn' phenomenon is found when the distant current electrode configuration of EIM is employed, but not when the current electrodes are placed close to the voltage measuring area. In that case the impedance qualitatively mimics the commonly used 3-element resistor–capacitor model. The possibility that the upturn is an artifact of the measuring system rather than a true property of the tissue is examined in detail. In particular, experiments are reported which argue against the transmission line mechanism proposed to explain similar increases in reactance in some high frequency whole-body BIA studies. Also, scaling of X versus R plots for muscle segments of different lengths strongly suggests that the upturn is as much a property of the underlying tissue as is the low frequency maximum in reactance.

Four-electrode impedance spectra of relaxed and contracted muscle biceps brachii were analyzed in an adult human subject over the frequency range from 300 Hz to 75 kHz. A feasibility of the principal component analysis of bioimpedance measurement for the evaluation of skeletal muscle contractile state was examined. The principal components score plots show a data grouping of the impedance spectra from the two muscle groups. The classification was performed using a soft independent modeling of class analogy (SIMCA) method. The data set comprised 32 samples (16 samples of contracted muscle and 16 samples of relaxed muscle). The leave-one-out test of the classification yields about 80% of correctly classified samples (11 samples for contracted and 15 samples for relaxed muscle).

In order to investigate the complex course of the electrical and mechanical processes of functional dyspepsia (FD), it is necessary to extract gastric motility information on both electricity and mechanism. According to the clinical standardization, 36 volunteers with functional dyspepsia were selected. The signal processing device has been designed by Chongqing University of Posts and Telecommunications. Multi-resolution analysis (MRA) decomposed the two signals of impedance gastric motility (IGM) and electrogastrogram (EGG) collected from the body surface. The wavelet transform is addressed to separate the IGM and EGG signals from impedance signals due to breathing and blood flow. By means of the energy and frequency spectrum analysis technique, the signals can be classified according to the dominant power and dominant frequency. Some indices, such as frequencies of EGG and IGM, signal power spectrum and dynamic spectrum, the rates of rhythm and power for the normal EGG and IGM and so on, can also be calculated. The primary experiments of gastric motility measurement and evaluation are executed by including healthy humans (control group: CG) and patients with FD (pathologic group: PG). There are significant differences in the temporal-domain and frequency-domain properties between the two groups. The main frequency of the CG belongs to 2–4 CPM and is clear and very regular, while the main frequency of the PG is much disordered. The peak of the maximal power of the CG belongs to 2–4 CPM and 1–2 CPM for the PG. The percentage of normal frequency (PNF) for the CG is 0.704 ± 0.255 and 0.402 ± 1.145 for the PG. The frequency instability coefficient (FIC) for the CG is 0.182 ± 0.059 and 0.374 ± 0.086 for the PG. The percentage of normal power (PNP) for the CG is 0.592 ± 0.044 and 0.468 ± 0.142 for the PG. The power instability coefficient (PIC) for the CG is 1.576 ± 0.481 and 4.006 ± 0.711 for the PG. The results of the experiments show that the proposed method of impedance can be a potential tool for the noninvasive assessment of gastric motility under gastrointestinal physiology and pathology conditions.

The time difference between the electrocardiogram and impedance cardiogram can be considered as a measure for the time delay between the electrical and mechanical activities of the heart. This time interval, characterized by the pre-ejection period (PEP), is related to the sympathetic autonomous nervous control of cardiac activity. PEP, however, is difficult to measure in practice. Therefore, a novel parameter, the initial systolic time interval (ISTI), is introduced to provide a more practical measure. The use of ISTI instead of PEP was evaluated in three groups: young healthy subjects, patients with Parkinson's disease, and a group of elderly, healthy subjects of comparable age. PEP and ISTI were studied under two conditions: at rest and after an exercise stimulus. Under both conditions, PEP and ISTI behaved largely similarly in the three groups and were significantly correlated. It is concluded that ISTI can be used as a substitute for PEP and, therefore, to evaluate autonomic neuropathy both in clinical and extramural settings. Measurement of ISTI can also be used to non-invasively monitor the electromechanical cardiac time interval, and the associated autonomic activity, under physiological circumstances.

A measurement technique for evaluation of processes occurring in the myocardium during blood flow arrest is presented. The system is devoted to monitoring the myocardium state during off-pump coronary artery by-pass grafting, i.e. during a surgical operation performed on a beating heart. A substantial part of the system is a probe whose construction is similar to a mechanical stabilizer. Such a stabilizer is used for preventing the myocardium from moving, thus it enables stitching of the graft and artery. In comparison to the commercially available stabilizer the developed probe is enhanced with a set of electrodes. These electrodes enable simultaneous measurements of impedance and electrograms. Examples of recordings made during in vivo studies are also presented. Thus, the proposed method is applicable and potentially very useful in clinical practice.

A multichannel logger for long-term measurements of sweat activity is presented. The logger uses skin surface electrodes for unipolar admittance measurements in the stratum corneum. The logger is developed with emphasis on clinical use. The portability of the logger enables recording of sweat activity under circumstances such as daily errands, exercise and sleep. Measurements have been done on 24 healthy volunteers during relaxation and exercise with heart rate monitoring. Recordings of sweat activity during sleep have been done on two healthy subjects. Early results show good agreement with the literature on sweating physiology and electrodermal activity. Results are presented showing measurements related to physical exercise, dermatomes, distribution of sweat glands and sympathetic activity. This study examines the normal sweating patterns for the healthy population, and we present results with the first 24 healthy volunteers. Comparing these results with similar measurements on hyperhidrosis patients will make it possible to find the most useful parameters for diagnosis and treatment evaluation.

Magnetic induction tomography is used to image the electrical properties inside a region of interest. The systems differ in the construction of the receiver channels which can be composed of coils or gradiometers. We will compare and discuss the image quality subject to two different types of receivers, different arrangements for the exciters and receivers and different signal-to-noise ratios. In order to evaluate the image quality, the point-spread function (PSF) was determined which is used for the calculation of the resolution and the correctness of the location of a perturbation. The results show that the PSF depends on (a) the location inside the object, (b) the type of receivers and (c) the configuration used, especially the location of the receiving and excitation channels. According to this, the local resolution is changed and has the maximum near the border of the object and decreases towards the centre of the object. In addition, the evaluation of the PSF shows a dislocation with respect to the underlying point-source position.

Magnetic induction tomography (MIT) has been suggested by several groups for the contact-less mapping of the passive electrical properties of tissues via AC magnetic fields in the frequency range between several tens of kHz and several tens of MHz. Multifrequency MIT as an analog to multifrequency EIT has been tried and first image reconstructions have been demonstrated with phantoms. MIT appears to yield comparable images to EIT but offers the advantage of being non-contacting. In the beta-dispersion range of most tissues the method is challenging because the signals are very small and buried in noise. In order to minimize drifts and systematic errors fast data acquisition is therefore pivotal. This paper presents a method for single-shot MIT which allows us to acquire the data for a multifrequency image with an analog bandwidth of 50 kHz–1.5 MHz which covers a good part of the β-dispersion of many tissues. The transmit (TX) coils are simultaneously driven by individual power amplifiers with a multisinus pattern with up to 3 A_pp. The amplifiers are configured as current sources so as not to perturb the excitation fields by inappropriately terminated coils. The separation of the different TX channels after reception is achieved by splitting up the carrier frequencies into individual subcarriers with a narrow spacing of at most 300 Hz. In this way every TX coil is identifiable by its own subcarrier but the whole excitation band is contained within a few kHz. The real and imaginary parts of the received signals are extracted efficiently with FFT. The system noise and the sources for low-frequency perturbations are analyzed and characterized.

Magnetic induction tomography (MIT) is a technique for imaging the internal conductivity distribution of an object. In MIT current-carrying coils are used to induce eddy currents in the object and the induced voltages are sensed with other coils. From these measurements, the internal conductivity distribution of the object can be reconstructed. In this paper, we introduce a 16-channel MIT measurement system that is capable of parallel readout of 16 receiver channels. The parallel measurements are carried out using high-quality audio sampling devices. Furthermore, approaches for reconstructing MIT images developed for the 16-channel MIT system are introduced. We consider low conductivity applications, conductivity less than 5 S m⁻¹, and we use a frequency of 10 MHz. In the image reconstruction, we use time-harmonic Maxwell's equation for the electric field. This equation is solved with the finite element method using edge elements and the images are reconstructed using a generalized Tikhonov regularization approach. Both difference and static image reconstruction approaches are considered. Results from simulations and real measurements collected with the Philips 16-channel MIT system are shown.

Progress made over the last few years in the hardware for magnetic induction tomography makes it possible to detect signal changes for low conductivity, low contrast variations in the target. However, image reconstruction of the data obtained presents a challenge especially when internal features within the target must be imaged. This paper focuses on the image reconstruction problem. A method of computing the forward problem with a lower computational cost than a model which needs to compute all the fields' values every time is presented. The method has been tested for its accuracy. Furthermore, the method can be used to produce a coefficient matrix which relates the contribution of each voxel to the total signal. This linear description of the problem forms the basis for image reconstruction. The images produced by direct linear reconstruction using the proposed matrix have been tested against maps produced by the dot product of the electric fields.

This paper presents a method for extrapolating the total body water (TBW) resistance R_t50 from the resistance measured at 50 kHz (R₅₀). A DXA examination and impedance measurements were carried out in a 1st group of 57 healthy volunteers with a Xitron 4200 multifrequency impedancemeter, in order to determine their values of R_t50 by comparison with resistances extrapolated at an infinite frequency by the Xitron (R_∞). TBW volumes were calculated using our modified BIS method (Jaffrin et al 2006 Med. Biol. Eng. Comput.44 873–82) from R_∞, R_t50 and from the fat-free mass measured by DXA, assuming a hydration rate of 73.2%. The same protocol and calculations were also carried out on a 2nd group of 21 subjects for independent validation. Data of the 1st group showed that values of R_t50, not significantly different from those of R_∞, could be obtained by dividing R₅₀ by 1.231 in men and by 1.224 in women. Applying this method to the 2nd group yielded also values of R_t50 not significantly different from R_∞. TBW volumes V_t50 obtained from R_t50 were not significantly different from those of our modified BIS method V_tn, or from TBW volumes obtained from DXA in both groups. A comparison with three BIA methods of TBW determination showed that our new method gave results in better agreement with TBW from DXA and from our modified BIS method.

The current paper proposes a new indicator for well-balanced dehydration of patients undergoing a haemodialysis. It is based on an estimator for the extra-cellular volume and the ultrafiltration rate. The extra-cellular fluid was computed from continuous tetrapolar bio-impedance measurements taken on the lower leg in a frequency range of several kilohertz up to 500 kHz. Finite element simulations on different leg models with anisotropic conductivities calculated with Cole models were carried out in order to incorporate the significant anisotropy of human tissue into the estimation process. The indicator was tested on measurement data gathered from 25 persons during 150 haemodialysis sessions. Its performance was determined by computing ROC curves. Results of the data analysis are reported.

Intradialytic fluid redistribution may cause hypotension. We hypothesized that measuring extracellular fluid volumes (ECV) with segmental bioimpedance analysis (SBIA) could test a simple, volume-driven model of redistribution among the arm, leg and trunk compartments. Patients (22, 5 females/17 males, with ages 60.2 ± 9 years, weights 80.7 ± 15 kg, heights 174 ± 9 cm) were studied during 30 HD treatments on different days. Hypotensive symptoms (Hypo+) were observed in eight patients. ECVs in the arm, trunk and leg, respectively V_A, V_T and V_L, were measured at initiation of, and throughout, dialysis. Two variables λ_A and λ_L were defined as V_A/V_T and V_L/V_T. System dynamics, assuming initial equilibrium, are then described by two rate coefficients k_RL and k_RA and two constants β and γ. These were obtained using a Marquardt–Levenberg least-squares algorithm. Significant differences (Hypo+ versus Hypo−) for λ_L (0.55 ± 0.13 versus 0.84 ± 0.3, *p < 0.05) and λ_A (0.17 ± 0.032 versus 0.23 ± 0.07, **p < 0.01) were found. The small value of λ_L might indicate that less leg volume predisposes to hypotension, larger peripheral volume mitigates hypotension. Observed transport ratios indicated that the ratio of limb to trunk volume stabilized during dialysis after an initial adjustment. These data imply encumbered movement of water from the interstitial components around skeletal muscle in the arm and leg to those of the trunk and are useful in predicting anatomical or situational predispositions to hypotension.

Although many methods have been utilized to measure degrees of body hydration, and in particular to estimate normal hydration states (dry weight, DW) in hemodialysis (HD) patients, no accurate methods are currently available for clinical use. Biochemcial measurements are not sufficiently precise and vena cava diameter estimation is impractical. Several bioimpedance methods have been suggested to provide information to estimate clinical hydration and nutritional status, such as phase angle measurement and ratio of body fluid compartment volumes to body weight. In this study, we present a calf bioimpedance spectroscopy (cBIS) technique to monitor calf resistance and resistivity continuously during HD. Attainment of DW is defined by two criteria: (1) the primary criterion is flattening of the change in the resistance curve during dialysis so that at DW little further change is observed and (2) normalized resistivity is in the range of observation of healthy subjects. Twenty maintenance HD patients (12 M/8 F) were studied on 220 occasions. After three baseline (BL) measurements, with patients at their DW prescribed on clinical grounds (DW_Clin), the target post-dialysis weight was gradually decreased in the course of several treatments until the two dry weight criteria outlined above were met (DW_cBIS). Post-dialysis weight was reduced from 78.3 ± 28 to 77.1 ± 27 kg (p < 0.01), normalized resistivity increased from 17.9 ± 3 to 19.1 ± 2.3 × 10⁻² Ω m³ kg⁻¹ (p < 0.01). The average coefficient of variation (CV) in three repeat measurements of DW_cBIS was 0.3 ± 0.2%. The results indicate that cBIS utilizing a dynamic technique continuously during dialysis is an accurate and precise approach to specific end points for the estimation of body hydration status. Since no current techniques have been developed to detect DW as precisely, it is suggested as a standard to be evaluated clinically.

The aim of this study was to evaluate the adequacy of single-frequency (sf-BIA) and multi-frequency bioelectrical impedance analyses (mf-BIA), in comparison with dual-energy x-ray absorptiometry (DXA), to evaluate body composition in maintenance haemodialysis (MHD) patients. Body composition of 27 adult MHD patients (9 f, 18 m), BMI 17.5–34.4 kg m⁻², was examined with DXA and BIA, with two different sf-BIA and 1 mf-BIA analysers. Biochemical markers of nutritional status and adequacy of dialytic treatment were also determined. Fat mass (FM) estimated by the different BIA analysers was found to be slightly but significantly higher than FM measured by DXA. In contrast, fat-free mass (FFM) obtained with BIA was found to be slightly but significantly lower than FFM DXA. No significant differences were found between LBM-DXA (that is FFM-DXA minus bone mass) and the different FFM BIA. The lowest mean prediction error versus DXA values was found with sf1BIA. In any case, a close correlation was found between all BIA values and DXA values, particularly for FFM. Furthermore, FFM and LBM results were significantly correlated with serum creatinine, which in MHD patients is an indicator of muscle mass. These results indicate that BIA can be used to evaluate body composition in MHD patients.