Table of contents

Participants at the First Mummy Range Workshop on Electrical Impedance Tomography at the Pingree Park of Colorado State University, 1--7 August 2002.

Electrical Impedance Tomography (EIT), the imaging of internal structure from external electric measurements, is a challenging problem. Designing and building high-precision measurement hardware for this error-sensitive modality requires state-of-the-art electrical engineering. The mathematical problem of reconstructing the image of unknown electrical parameter distributions from measured data is very complicated since the problem is nonlinear and ill-posed. And finally, the interpretation of the resulting images in medical, geophysical or industrial applications is not yet fully understood and needs application-oriented research.

The First Mummy Range Workshop on Electrical Impedance Tomography (see www.eitworkshop.org) was held during 1–7 August 2002, at the Pingree Park of Colorado State University at an altitude of 9000 feet. The monumental Rocky Mountains created an exhilarating atmosphere for the 60 participants of the meeting. This Workshop was part of a series of meetings in EIT, the first of which was held in Sheffield, UK, in 1986 under the sponsorship of the European Community. Since then meetings have been held nearly annually in the UK, but this was the first to be held in the US. Papers from many of the previous meetings, including those held during 1999, 2000 and 2001, have been published in special issues of Physiological Measurement.

The mixed audience in the Mummy Range Workshop consisted of experts in electrical engineering, mathematics, physics, medicine and geophysics. This allowed the exchange of ideas across the boundaries of traditional fields of study, which resulted in the broad range of topics presented, and a different atmosphere from traditional conferences for a single academic field. The diversity amongst the Workshop participants is evident in the collection of papers on pages 391--638 of this special issue. All three viewpoints on EIT mentioned above are covered. New contributions to hardware design are presented. Studies on the reconstruction problem include new algorithms and examples of anisotropic conductivity distributions that cannot be detected by EIT. Clinical applications to human head imaging and imaging of pulmonary perfusion are addressed in this issue, and electrical properties of tissues relevant for EIT imaging are discussed. Another significant feature in this collection is that there are five papers addressing new EIT technology, three of them on MREIT, a new technique combining EIT with MRI.

The advantages of an interdisciplinary workshop were probably most evident in the evening discussion sessions in Pingree Park's fireplace lounge. Three discussion sessions were held addressing issues of hardware, reconstruction algorithms, and the future of EIT. These were well-attended and provoked lively debate on issues such as 2D versus 3D reconstruction algorithms and the effects of electrode position and the precision of electrode placement. The discussion session on the future of EIT included a serious discussion about the clinical applications of the field. Many applications were discussed including the detection of breast cancer, the location of the focus of an epiliptic seizure, the imaging of action potentials in the brain, monitoring cardiac function, monitoring mechanical ventilation, detection of pulmonary embolus, and high altitude pulmonary edema. Challenges were also brought up by the physicians present, such as the possibility of using no electrodes whatsoever for EIT and the possibility of grafting EIT on an existing apparatus such as an ultrasound probe or mammography. It was agreed that EIT's greatest potential lies in its portability and convenience as a bedside monitoring device. It has a unique advantage when patients cannot remain still, such as is the case with neonates or epileptic seizure patients.

The future of EIT looks very healthy indeed, and the collection of papers in this special issue provides further evidence of the significant advances that are being made in all aspects of the field. We thank the participants of the First Mummy Range Workshop, the authors of the papers published herein, and the many reviewers of those papers, for their contributions to the success of this special issue of Physiological Measurement

David Isaacson  Department of Mathematics, Rensselaer Polytechnic Institute, Troy, NY 12180, USAJennifer Mueller  Department of Mathematics, Colorado State University, Fort Collins, CO 80525, USASamuli Siltanen  Imaging Division, Instrumentarium Corporation, Tuusula, FIN-04301, Finland Guest Editors

A simple pharmacokinetic model to explain the time course of [O-15]water in human whole blood after bolus injection is described. The model has been derived from measurements in twelve healthy volunteers who were measured repeatedly, resulting in 67 datasets, made in the context of PET blood flow studies. In contrast to traditional volume of distribution estimates of total body water (TBW) which rely on measurements after many hours, the model and data provide insights into the fast uptake components in the distribution of water in the body. Data fitting shows that the volume of distribution of fast exchanging tissues is 21 l, TBW was calculated to be 37 l. Monte Carlo simulation showed that the expected inaccuracy of determination of parameters due to unsystematic sources in the measurement data was around 5% for most parameters. Our data show that water extraction to tissue is somewhat higher than would be predicted from the tabulated values, probably because skeletal blood flow is sensitive to physiological status and environmental conditions. The study provides valuable reference data on the distribution and kinetics of water in man. Using the parameters and model from this study, reference input time-activity curves can be calculated, e.g. for the Monte Carlo study of error propagation in PET studies.

Knowledge of electrical tissue conductivity is necessary to determine deposition of electromagnetic energy and can further be used to diagnostically differentiate between normal and neoplastic tissue. We measured 17 rats with a total of 24 tumours of the K12/TRb rat colon cancer cell line. In each animal we measured in vivo hepatic tumour and normal tissue conductivity at seven frequencies from 10 Hz to 1 MHz, at different tumour stages between 6 and 12 weeks after induction. Conductivity of normal liver tissue was 1.26 ± 0.15 mS cm⁻¹ at 10 Hz, and 4.61 ± 0.42 mS cm⁻¹ at 1 MHz. Conductivity of tumour was 2.69 ± 0.91 mS cm⁻¹ at 10 Hz, and 5.23 ± 0.82 mS cm⁻¹ at 1 MHz. Conductivity was significantly different between normal and tumour tissue (p < 0.05). We determined the percentage of necrosis and fibrosis at the measurement site. We fitted the conductivity data to the Cole–Cole model. For the tumour data we determined Spearman's correlation coefficients between the Cole–Cole parameters and age, necrosis, fibrosis and tumour volume and found significant correlation between necrosis and the Cole–Cole parameters (p < 0.05). We conclude that necrosis within the tumour and the associated membrane breakdown is likely responsible for the observed change in conductivity.

This paper presents a mathematical model of the oxygen alveolo–capillary exchange to provide the capillary oxygen partial pressure profile in normal and pathological conditions. In fact, a thickening of the blood–gas barrier, heavy exercise or a low oxygen partial pressure (PO₂) in the alveolar space can reduce the O₂ alveolo–capillary exchange. Since the reversible binding between haemoglobin and oxygen makes it impossible to determine the closed form for the mathematical description of the PO₂ profile along the pulmonary capillaries, an approximate analytical solution of the capillary PO₂ profile is proposed. Simulation results are compared with the capillary PO₂ profile obtained by numerical integration and by a piecewise linear interpolation of the oxyhaemoglobin dissociation curve. Finally, the proposed model is evaluated in a large range of physiopathological diffusive conditions. The good fit to numerical solutions in all experimental conditions seems to represent a substantial improvement with respect to the approach based on a linear approximation of the oxyhaemoglobin dissociation curve, and makes this model a candidate to be incorporated into the integrated descriptions of the entire respiratory system, where the datum of primary interest is the value of end capillary PO₂.

This work evaluates the feasibility of monitoring ischemic injury in the gastrointestinal mucosa by impedance spectroscopy, using a minimally invasive intestinal catheter. The disruption of the intestinal mucosa plays a key role in the evolution of shock and is the 'motor of multiple organ failure'. Different technologies have been developed to monitor mucosal perfusion, oxygenation and/or ischemia, but no practical method exists to assess tissue damage, which may be crucial for preventing multiple organ failure. The experimental protocol of this study relied on an isobaric model of hypovolemic shock in 16 anaesthetized rabbits assigned to three groups: sham (n = 6), ischemia (n = 5) and ischemia + reperfusion (n = 5). Complex impedance spectra were recorded in the range of 0.05 to 300 kHz, with simultaneous measurements of tonometric pHi in the ileum every 30 min for 4 h. Impedance spectra were reproducible, and those of tissue under prolonged ischemia were clearly differentiable from those of normally perfused tissue. The dynamic changes in impedance did not correlate directly with either tissue perfusion or pHi, but instead correlated well with the duration of ischemia. It is concluded that impedance spectroscopy does indeed measure changes in tissue injury, and could be a very useful tool to guide therapy of patients in shock.

Biological tissues undergoing inflammation and dysplasia seem to exhibit changes in the intercellular space that can be sensed using low frequency electrical impedance methods. Basically, low frequency electric current flows through this space and its widening as well as the disruption of the tight junction decrease the resistance, facilitating current flow. The electrical changes accompanying structural changes from columnar tissue to adenocarcinoma in Barrett's metaplastic mucosa and gastric tissue are illustrated using resected tissue from 32 patients. Two hundred and fifty-eight biopsies were analysed, correlating their electrical resistivity (R) at 9.6 kHz and their histopathological interpretation. Compared to non-inflamed non-dysplastic columnar tissue (R = 4.9 Ω m), the results suggest a small but statistically significant decrease of electrical impedance in columnar tissue showing inflammation (R = 4.2 Ω m, p = 0.016) and a larger decrease when dysplasia is present (R = 3.4 Ω m, p = 0.040). If this method is validated further, this technique could be used to obtain guided biopsies from patients undergoing surveillance programmes for Barrett's oesophagus. We aim to refine this technique using a new system with lower frequencies and, possibly, in vitro (cultured cells) and in vivo (rats) models of Barrett's oesophagus.

It is accepted that older subjects have increasing arterial stiffness resulting in changes in the propagation of the pulse to the periphery, and thereby influencing the peripheral pulse timing and shape characteristics. However, this age association with pulse shape is less clear in younger subjects and for different peripheral measurement sites. The aim of this study was to determine the association between age and changes in pulse shape characteristics at the ears, fingers and toes. Photoplethysmography pulse waveforms were recorded non-invasively from the right and left sides at the ears, index fingers and great toes of 116 normal healthy human subjects. Their median age was 41 years (range 13–72) allowing four distinct age groups to be considered (subjects younger than 30 years, 30–39 years, 40–49 years and 50 years of age or older). Normalized ear, finger and toe pulse shapes were calculated, for the whole subject group, and for the subjects within each age group. The differences in shape, relative to the oldest group, were also calculated for two distinct regions of interest; the systolic rising edge and the dicrotic notch of the pulse. Subtle, gradual and significant changes in the pulse shape were found at all sites with overall elongation of the systolic rising edge (p < 0.05) and damping of the dicrotic notch (p < 0.05) with age. The overall age-related changes in multi-site PPG pulse shape characteristics at the ear, finger and toe sites have been demonstrated and quantified. Age-matched normal ranges must be considered when evaluating pulses from patients with possible vascular disease.

The purpose of this research is to select the best features to have a high rate of motion classification for controlling an artificial hand. Here, 19 EMG signal features have been taken into account. Some of the features suggested in this study include combining wavelet transform with other signal processing techniques. An assessment is performed with respect to three points of view: (i) classification of motions, (ii) noise tolerance and (iii) calculation complexity. The energy of wavelet coefficients of EMG signals in nine scales, and the cepstrum coefficients were found to produce the best features in these views.

This work describes the fundamentals and calibration procedure of an instrument for in vivo evaluation of the heat convection coefficient between the endocardium and the circulating blood flow. The instrument is to be used immediately before radio-frequency cardiac ablation is performed. Thus, this instrument provides researchers with a valuable parameter to predict lesion size to be achieved by the procedure. The probe is a thermistor mounted in a Swan–Ganz catheter, and it is driven by a constant-temperature anemometer circuit. A 1D model of the sensor behaviour in a convective medium, the calibration procedure and the apparatus are explained in detail. Finally, a performance analysis of the instrument in the range of 200–3500 W m⁻² K⁻¹ shows that the average absolute error of full scale is 7.4%.

The mechanical nature of gastric contraction activity (GCA) plays an important role in gastrointestinal motility. The aim of this study was to detect GCA in anaesthetized dogs, using simultaneously the techniques of AC biosusceptometry (ACB) and manometry, analysing the characteristics of frequency and amplitude (motility index) of GCA, modified by drugs such as prostigmine and N-butyl-scopolamine. The ACB method is based on a differential transformer of magnetic flux and the magnetic tracer works as a changeable external nucleus. This magnetic tracer causes a modification in the magnetic flux, which is detected by the coils. The results obtained from the ACB showed a performance comparable to the manometry in measuring the modifications in the frequency and amplitude of the GCA. We concluded that this ACB technique, non-invasive and free of ionizing radiation, is an option for evaluating GCA and can be employed in future clinical studies.

Stroke is a major cause of disability within the western world. About 20% of strokes are a consequence of atheromatous narrowing of the origin of the internal carotid artery. Carotid endarterectomy has been shown to be an effective treatment for those with symptomatic and severe stenosis, provided the risk of death and peri-operative stroke is less than 7%. The aim of this study was to investigate the clinical value of jugular venous oxygen saturation (SJVO2) monitoring in identifying patients who develop cerebral ischaemia whilst undergoing an awake carotid endarterectomy by comparison with a simple neurological assessment. Each patient underwent a standard awake carotid endarterectomy. Per-operatively a SJVO2 catheter was inserted, and the jugular oxygen saturation was correlated with the presence or absence of cerebral ischaemia. Data from 34 patients were analysed using time-series plots and by calculating a receiver operator characteristic (ROC) curve. The optimal sensitivity and specificity for this technique were found to be 1.0 and 0.8, respectively, when a 25% change in SJVO2 was used as a threshold. Although a small observational study, we have shown that percentage change in SJVO2 correlates well with the development of clinically apparent cerebral ischaemia. This technique may improve the safety of carotid endarterectomy under general anaesthesia when used with other more established monitoring methods.

Ultrasonic transit-time airflow meters (UFM) allow simultaneous measurements of volume flow V'(t) and molar mass MM(t) of the breathing gas in the mainstream. Consequently, by using a suitable tracer gas the functional residual capacity (FRC) of the lungs can be measured by a gas wash-in/wash-out technique. The aim of this study was to investigate the in vitro accuracy of a multiple-breath wash-in/wash-out technique for FRC measurements using 4% sulphur hexafluoride (SF₆) in air. V'(t) and MM(t) were measured with a Spiroson SCIENTIFIC flowmeter (ECO Medics, CH) with 1.3 ml dead space. Linearity of airflow and MM were tested using different tidal volumes (V_T and breathing gases with different O₂ and SF₆ concentrations. To determine the accuracy of FRC measurements SF₆ wash-in and wash-out curves from four mechanical lung models (FRC of 22, 53, 102 and 153 ml) were evaluated by the Spiroson. For each model five measurements were performed with a physiological V_T/FRC ratio of 0.3 and constant respiratory rate of 30 min⁻¹. The error of measured V_T (range 4–60 ml) was <2.5%. There was a strong correlation between the measured and calculated MM of different breathing gases (r = 0.989), and the measuring accuracy was better than 1%. The measured FRC of the four models were 20.3, 49.7, 104.3 and 153.4 ml with a coefficient of variation of 16.5%, 4.5%, 4.9% and 3%. Accordingly, for FRC < 100 ml the in vitro accuracy was better than 8% and for FRC > 100 ml better than 2.5%. The determination of FRC by MM measurements using the UFM is a simple and cost-effective alternative to conventionally used gas analysers with an acceptable accuracy for many clinical purposes.

The intra- and inter-subject variabilities of the cerebral dynamic autoregulatory index (ARI) were studied in a group of 14 healthy subjects aged 23–51 years. An alternative index, derived from autoregressive-moving average (ARMA) modelling of the arterial blood pressure (ABP)–cerebral blood flow velocity (CBFV) dynamic relationship, named ARMA-ARI, is also proposed. The susceptibility of both indices to physiological sources of variability was studied by performing measurements during spontaneous respiration (SR), and controlled breathing at 6, 10 and 15 breaths min⁻¹. ABP was measured non-invasively (Finapres), CBFV was recorded with Doppler ultrasound in both middle cerebral arteries and end-tidal CO₂ (EtCO₂) was estimated with an infrared capnograph. ARI and ARMA-ARI were calculated as a summary measure for the whole of each recording period, and also continuously, using a 60 s moving data window. Respiration did not have an effect on either of these indices, despite significant, but relatively small, reductions in EtCO₂ at 10 and 15 bpm, compared to SR. Very significant differences were observed between ARI and ARMA-ARI in relation to their stability, variability and sensitivity to discriminate between subjects. For continuous estimates the coefficient of variation of ARI was 30 ± 21% compared to 15 ± 8% for ARMA-ARI (p < 0.000). The cumulative probability distributions were also significantly different for the two indices for each of the respiratory manoeuvres. The greater stability and reduced variability of ARMA-ARI, in relation to the classic ARI, suggest that the former should be used in future studies of dynamic autoregulation, mainly in situations where an improved temporal resolution might be required, such as the investigation of vaso-vagal syncope or the physiology of exercise.

The ability to objectively determine the degree of tissue edema and to monitor on-line fluid balance in critically ill patients would be a clinical benefit. In this prospective descriptive trial, we evaluated a new noninvasive method—dielectric constant of skin and subcutaneous fat (SSF)—in assessing fluid balance during cardiac surgery. The dielectric constant at the applied high radiofrequency is a direct measure of tissue water content. Twenty-nine patients with elective cardiac surgery participated in the study. Dielectric constants on forearm, thigh and abdomen were measured before surgery, within 1 h after surgery and in the first, second, third and fourth postoperative morning. At the same time the patients were weighed, except immediately after the operation and the first postoperative day when fluid balances were calculated. A statistically significant correlation (r = 0.60, p < 0.01) was found between the increase of the dielectric constant of SSF and weight gain of the patients from the baseline to the second postoperative morning. From the second to the fourth postoperative day when the patients were losing the weight, a statistical significant correlation between the dielectric constant and weight loss was not found. The results suggest that the measurement of the dielectric constant is a promising new method in assessing the fluid status of operated patients during the time the patients cannot be weighed.

In this paper we describe a new, direct and mathematically exact method for the reconstruction of the isotropic conductivity in a plane body from static electric measurements on the boundary of the body. The method is inspired by a uniqueness proof for the inverse conductivity problem due to Brown–Uhlmann and covers conductivities having essentially one derivative. Moreover, we give a numerical implementation of the algorithm and test the performance on a simple, synthetic example.

Electrical impedance tomography (EIT) is a non-invasive technique used to image the electrical conductivity and permittivity within a body from measurements taken on the body surface. Four methods are being investigated for breast cancer diagnosis by EIT today: Single voltage source, single current source and multiple current sources with a fixed pre-determined 'canonical' pattern of currents and an adaptively determined 'optimal' pattern of currents. To determine which of these four methods might yield the best distinguishability using planar electrode arrays for breast cancer detection, we placed electrode arrays on a saline tank and used each excitation pattern to detect a conducting target placed at the centre of a flat electrode array in two geometries: mammography geometry and single probe geometry. The result was that the multiple current sources method had higher distinguishability than either the SCS or the SVS method. In both these electrode geometries, the optimal current pattern had higher distinguishability than the other patterns at all distances.

We construct anisotropic conductivities in dimension 3 that give rise to the same voltage and current measurements at the boundary of a body as a homogeneous isotropic conductivity. These conductivities are non-zero, but degenerate close to a surface inside the body.

In electrical impedance tomography surface measurements of voltages and currents are recorded and the image reconstruction algorithm uses this set of boundary data to estimate internal electrical properties of the region under investigation. Therefore correct and accurate modelling of the current and voltage distributions (forward model) is an essential part of any reconstruction method. In this paper, we explored the root cause of a boundary layer effect in the reconstructed conductivity map and found it to be an artefact arising from 2D to 3D data-model mismatch within the imaging algorithm. We propose a data calibration scheme that improves the reconstruction results by removing these boundary or edge effects. We present both two-dimensional and three-dimensional images for agar phantoms using this data calibration scheme which are markedly better than their counterparts recovered when the measurement data are not calibrated with the procedure outlined herein.

A major drawback of electrical impedance tomography is the poor quality of the conductivity images, i.e., the low spatial resolution as well as large errors in the reconstructed conductivity values. The main reason is the necessity for regularization of the ill-conditioned inverse problem which results in excessive spatial low-pass filtering.

A novel regularization method (SMORR (spectral modelling regularized reconstructor)) is proposed, which is based on the inclusion of spectral a priori information in the form of appropriate tissue models (e.g. Cole models). This approach reduces the ill-posedness of the inverse problem, when multifrequency data are available. An additional advantage is the direct reconstruction of the (physiological) tissue parameters of interest instead of the conductivities.

SMORR was compared with posterior fitting of a Cole model to the conductivity spectra obtained with a classical iterative reconstruction scheme at various frequencies. SMORR performed significantly better than the reference method concerning robustness against noise in the data.

This paper presents a new application of a generalized vector sample pattern matching (GVSPM) method for image reconstruction of conductivity changes in electrical impedance tomography. GVSPM is an iterative method for linear inverse problems. The key concept of the GVSPM is that the objective function is defined in terms of an angular component between the inner product of the known vector and solution of a system of equations. Comparisons are presented between images of simulated and experimental data, reconstructed using truncated singular value decomposition and GVSPM. In both cases, a normalized sensitivity matrix is constructed using the finite volume method to solve the forward problem.

A finite difference model of the human thorax with 113 400 control volumes (nodes) based on ECG gated MRI images was used to evaluate the Sheffield DAS-01P EIT system. Sixteen simulated electrode positions equally spaced around the thorax model at approximately the fourth intercostals space level were selected. Pairs of adjacent positions were excited sequentially by injecting current in a manner similar to that used by the Sheffield DAS-01 P EIT system. The resulting voltages on the non-excited electrode positions were calculated and used to reconstruct the image using the Sheffield filtered back projection algorithm. By changing the resistivities of the lungs, the ventricles and the atria over a range of 1% to 40%, the resulting changes in the images were quantified by measuring the average resistivity change over a region defined automatically by two thresholds, 40% or 80% of the average of the first four pixels with the largest change. The results show that the changes observed in the images are consistently less than the changes in the model, but changed in a nearly linear manner as a function of resistivity in the model. For 40% resistivity changes in the model for right lung, right ventricle and right atrium, the observed resistivity changes in the region of interest (ROI, defined by the 80% threshold) of the images are 32% for the right lung, 11% for the right ventricle and 5.5% for the right atrium, which suggests strong volume dependence of EIT imaging. The effect of structural (size) change between end diastole and end systole was also studied, which showed large resistivity changes caused in the heart region of the constructed image. The study demonstrates that the Sheffield DAS-01P EIT reconstruction algorithm tracks the change occurring in the lungs most closely and with proper scaling may be used to observe physiological changes.

Electrical impedance tomography (EIT) is a recently developed medical imaging method which has practical advantages for imaging brain function as it is inexpensive, rapid and portable. Its principal use in validated human studies to date has been to image changes in impedance at a single excitation frequency over time, but there are potential applications where it is desirable to obtain images from a single point in time, which could be achieved by imaging over multiple frequencies. We describe a novel multifrequency EIT design which provides up to 64 electrodes for imaging in the head. This was achieved by adding a multiplexer to a single channel of an existing system, the Sheffield Mark 3.5. This provides a flexible protocol for addressing up to 64 electrodes but CMRR decreases from 90 dB to 80 dB and analogue amplifier bandwidth from >1.6 MHz to 0.8 MHz. This did not significantly affect performance, as cylinders of banana, 10% of the diameter of a saline filled spherical tank, could be visualized with frequency referenced imaging. The design appears to have been an acceptable compromise between practicality and performance and will now be employed in clinical trials of multifrequency EIT in stroke, epilepsy and neonatal brain injury.

A possible clinical application of electrical impedance tomography (EIT) might be to monitor changes in the pulmonary circulation, provided the reproducibility of the EIT measurement is adequate. The purpose of this study was threefold: the intra- and inter-investigator variability of repeated measurements was investigated. Three different regions of interest (ROI) were analysed to assess the optimal ROI. Twenty-four healthy subjects and six patients were included. The Sheffield applied potential tomograph (DAS-01P, IBEES, Sheffield, UK) was used. Electrodes were attached by investigator A, and duplicate EIT measurements were performed. After detachment and 45 min of rest, the protocol was repeated by another investigator B, and afterwards by the initial investigator A. Three ROIs were analysed: whole circle, 'inner half circle' and contour. The mean difference in impedance changes between observers is presented in arbitrary units (AU) ± SD. Finally, the influence of age, body composition and sex on the EIT result was examined. For the contour ROI, the mean difference for the intra-investigator situation was −1.44 × 10⁻² ± 18.45 × 10⁻² AU (−0.7 ± 9.0%), and was 5.46 × 10⁻² ± 21.66 × 10⁻² AU (2.7 ± 10.8%) for the inter-investigator situation. The coefficient of reproducibility of the intra- and inter-investigator reproducibility varied between 0.89 and 0.97 for all ROIs (P < 0.0001). There is a relation between impedance change and age (correlation coefficient r = −0.63, P < 0.01 for contour ROI), and between impedance change and body mass index (BMI) (r = −0.53, P < 0.05). We found a significant difference in mean impedance change between groups of males and females. In conclusion, EIT results are highly reproducible when performed by the same investigator as well as by two different investigators.

A single-channel MIT measuring system for obtaining phase delays is given. The circuit, which is described in detail, uses a high-frequency analogue multiplier to measure the phase difference between the signal and a reference signal. The noise in the phase measurement is ∼1.5 millidegree when the time constant of measuring is 0.1 s, and the drift over about 1 day is ∼10 millidegree.

Questions regarding the feasibility of using electrical impedance tomography (EIT) to detect breast cancer may be answered by building a sufficiently precise multiple frequency EIT instrument. Current sources are desirable for this application, yet no current source designs have been reported that have the required precision at the multiple frequencies needed. We have designed an EIT current source using an enhanced Howland topology in parallel with a generalized impedance converter (GIC). This combination allows for nearly independent adjustment of output resistance and output capacitance, resulting in simulated output impedances in excess of 2 GΩ between 100 Hz and 1 MHz. In this paper, the theoretical operation of this current source is explained, and experimental results demonstrate the feasibility of creating a high precision, multiple frequency, capacitance compensated current source for EIT applications.

Tetrapolar probes have been widely used for measuring the impedance spectra of tissues. However, the non-uniform sensitivity distribution of these probes limits the ability to identify conductivity changes in tissue. This paper presents a novel method for improving the sensitivity distribution beneath a tetrapolar probe. The method consists of placing a hydrogel layer between the probe and the tissue in order to make the sensitivity positive everywhere within the tissue. Theoretical and measured sensitivity distributions are compared. A good agreement between theoretical and measured data from an electrolytic tank was obtained with a maximum error of 1.3%. In vivo forearm measurements showed that the use of a conductive layer does enable tissue conductivity spectra to be determined. A smaller variation between subjects was obtained when using the stand-off. It was not possible to assess the absolute accuracy of the method due to the absence of a 'gold standard' for the measurement of tissue conductivity spectra.

Three types of commercially available headnet electrode arrays, designed for use in EEG, and conventional EEG Ag/AgCl cup electrodes were tested on human subjects, and a realistic, saline-filled head-shaped tank was prepared with vegetable skin to simulate human skin in order to determine the optimum electrode system for electrical impedance tomography (EIT) of the human head. Impedance changes during EIT acquisition were produced in healthy volunteers during a finger–thumb apposition task and in tanks by the insertion of a Perspex rod. Signal-to-baseline noise, measured from raw EIT data, was 2.3 ± 0.3 and 2.3 ± 0.2 for the human and tank data, respectively. In both the human and tank experiments, a commercial hydrogel elasticated electrode headnet produced the least amount of baseline noise, and was the only headnet in the human data with noise levels acceptable for EIT imaging. Image quality measured in the tank was similar for most of the headnets tested, except that the EEG electrodes produced a higher positional error and electrodes in a geodesic elasticated net produced images with worse subjective image quality. Overall, the hydrogel elasticated headnet was judged to be the most suitable for human neuroimaging with EIT.

Magnetic induction tomography (MIT) is used for reconstructing the changes of the conductivity in a target object using alternating magnetic fields. Applications include, for example, the non-invasive monitoring of oedema in the human brain. A powerful software package has been developed which makes it possible to generate a finite element (FE) model of complex structures and to calculate the eddy currents in the object under investigation. To validate our software a model of a previously published experimental arrangement was generated. The model consists of a coaxial coil system and a conducting sphere which is moved perpendicular to the coil axis (a) in an empty space and (b) in a saline-filled cylindrical tank. The agreement of the measured and simulated data is very good when taking into consideration the systematic measurement errors in case (b). Thus the applicability of the simulation algorithm for two-compartment systems has been demonstrated even in the case of low conductivities and weak contrast. This can be considered an important step towards the solution of the inverse problem of MIT.

In magnetic induction tomography (MIT) the in-quadrature component, and hence the phase, of the received signal contains information about the conductivity of the tissue. The quality of imaging will depend on the precision with which phase can be measured. Preliminary studies suggest that a precision of 10 m° may be required for a practical biomedical MIT system operating at 10 MHz. This paper describes the results of measurements carried out with a 16-channel, downconverting, 10 MHz, MIT system utilizing two types of data extraction techniques: direct-phase measurement and measurement of the in-phase and in-quadrature components of the signal with a vector voltmeter. The basic precision provided by each technique was 50 m°, with thermal drift representing the major limiting factor. Preliminary measurements of average conductivity and permittivity for a human thigh in vivo are given.

In magnetic resonance current density imaging (MRCDI), we inject current into a subject through surface electrodes and measure the induced magnetic flux density B inside the subject using an MRI scanner. Once we have obtained all three components of B, we can reconstruct the internal current density distribution J = ∇ × B/μ₀. This technique, however, requires subject rotation since the MRI scanner can measure only one component of B that is parallel to the direction of its main magnetic field. In this paper, under the assumption that the out-of-plane current density J_z is negligible in an imaging slice belonging to the xy-plane, we developed an imaging technique of current density distributions using only B_z, the z-component of B. The technique described in this paper does not require a subject rotation but the quality of reconstructed images depends on the amount of out-of-plane current density J_z. From numerical simulations, we found that the new algorithm could be applied to subjects such as human limbs using longitudinal electrodes.

In magnetic resonance electrical impedance tomography (MREIT) we inject currents through electrodes placed on the surface of a subject and try to reconstruct cross-sectional resistivity (or conductivity) images using internal magnetic flux density as well as boundary voltage measurements. In this paper we present a static resistivity image of a cubic saline phantom (50 × 50 × 50 mm³) containing a cylindrical sausage object with an average resistivity value of 123.7 Ω cm. Our current MREIT system is based on an experimental 0.3 T MRI scanner and a current injection apparatus. We captured MR phase images of the phantom while injecting currents of 28 mA through two pairs of surface electrodes. We computed current density images from magnetic flux density images that are proportional to the MR phase images. From the current density images and boundary voltage data we reconstructed a cross-sectional resistivity image within a central region of 38.5 × 38.5 mm² at the middle of the phantom using the J-substitution algorithm. The spatial resolution of the reconstructed image was 64 × 64 and the reconstructed average resistivity of the sausage was 117.7 Ω cm. Even though the error in the reconstructed average resistivity value was small, the relative L²-error of the reconstructed image was 25.5% due to the noise in measured MR phase images. We expect improvements in the accuracy by utilizing an MRI scanner with higher SNR and increasing the size of voxels scarifying the spatial resolution.

Magnetic resonance–electrical impedance tomography (MR–EIT) was first proposed in 1992. Since then various reconstruction algorithms have been suggested and applied. These algorithms use peripheral voltage measurements and internal current density measurements in different combinations. In this study the problem of MR–EIT is treated as a hyperbolic system of first-order partial differential equations, and three numerical methods are proposed for its solution. This approach is not utilized in any of the algorithms proposed earlier. The numerical solution methods are integration along equipotential surfaces (method of characteristics), integration on a Cartesian grid, and inversion of a system matrix derived by a finite difference formulation. It is shown that if some uniqueness conditions are satisfied, then using at least two injected current patterns, resistivity can be reconstructed apart from a multiplicative constant. This constant can then be identified using a single voltage measurement. The methods proposed are direct, non-iterative, and valid and feasible for 3D reconstructions. They can also be used to easily obtain slice and field-of-view images from a 3D object. 2D simulations are made to illustrate the performance of the algorithms.

Impedance measurement is a promising technique for detecting pre-malignant changes in epithelial tissue. This paper considers how the design of the impedance probe affects the ability to discriminate between tissue types. To do this, finite element models of the electrical properties of squamous and glandular columnar epithelia have been used. The glandular tissue model is described here for the first time. Glandular mucosa is found in many regions of the gastrointestinal tract, such as the stomach and intestine, and has a large effective surface area. Firstly, the electrical properties of a small section of gland, with epithelial cells and supportive tissue, are determined. These properties are then used to build up a three-dimensional model of a whole section of mucosa containing many thousands of glands. Measurements using different types of impedance probe were simulated by applying different boundary conditions to the models. Transepithelial impedance, and tetrapolar measurement with a probe placed on the tissue surface have been modelled. In the latter case, the impedance can be affected by conductive fluid, such as mucus, on the tissue surface. This effect has been investigated, and a new design of probe, which uses a guard electrode to counteract this potential source of variability, is proposed.

Electrical bioimpedance spectroscopy is a fast and relatively easily applicable method for tissue characterization. In the frequency range up to 10 MHz, current conduction through tissue is mainly determined by tissue structure, i.e. the extra- and intra-cellular compartments and the insulating cell membranes. Therefore, changes in the extra- and intra-cellular fluid volumes are reflected in the impedance spectra. Investigations of tumours (DS sarcoma, implanted on the hind foot dorsum of rats) during treatment with localized hyperthermia (HT), photodynamic therapy (PDT) and the combination of these two components were carried out using impedance spectroscopy in the frequency range of 37 Hz to 3.7 MHz. Data collected reveal totally different, but characteristic, behaviour patterns for the three treatments. HT caused a slow increase in conductance at high frequencies (G_HF) and in the extracellular space index (ECSI), indicating an increase in extracellular fluid volume and in total fluid content. With PDT, G_HF increased immediately upon commencement of irradiation and was accompanied by a distinct decrease in ECSI, indicating the development of a pronounced intracellular oedema.

Send notice of meetings to: Dr S J Meldrum, Norfolk and Norwich Hospital, Norwich NR1 3SR, UK

2003

10th Congress of the World Federation for Ultrasound in Medicine and Biology 1 - 4 June, Montreal, Cananda www.aium.org

AAMI 2003 -- Annual Meeting of the Association for the Advancement of Medical Instrumentation 14 - 17 June, Long Beach, CA, USA www.aami.org/meetings/aami2003

CARS 2003 -- 17th International Congress and Exhibition on Computer Assisted Radiology and Surgery 25 - 28 June, London, UK www.cars-int.de

Physics and Engineering of Blood Flow 3 July 2002, York, UK http://www.ipem.org.uk/meetings/2003.html

7th Annual Meeting on Medical Image Understanding and Analysis 10 - 11 July, Sheffield, UK www.sys.uea.ac.uk/~rz/miua2003/

40th Annual Conference of the American Society for Healthcare Engineering 14 - 16 July, San Antonio, TX, USA www.hospitalconnect.com/ashe/conference/welcome.html

Fifth International Conferences on Neural Networks and Expert Systems in Medicine and Healthcare 21 - 23 July, Sheffield, UK www.shu.ac.uk/conference/nnesmed/index.html

Medical Imaging Analysis: Modalities Using Non-Ionizing Electromagnetic Radiation 10 – 11 August, San Diego, USA http://www.aapm.org/meetings/03IS/category_descriptions.asp

5th IFAC Symposium on Modelling and Control in Biomedical Systems 21 - 23 August, Melbourne, Australia www.tourhosts.com.au/ifac2003

World Congress of Medical Physics and Bioengineering 24 - 29 August, Sydney, Australia www.eng.unsw.edu.au/wc2003

Annual Congress of the European Society of Cardiology 31 August - 3 September, Vienna, Austria www.escardio.org

Annual Scientific Meeting of the Institute of Physics and Engineering in Medicine 15 - 17 September, Bath, UK http://www.ipem.org.uk/meetings/ASM2003.html

25th Annual International Conference of the IEEE Engineering in Medicine and Biology Society 17 - 21 September, Cancun, Mexico http://itzamna.uam.mx/cancun

International Conference on Biomedical Optics 19 - 23 September, Bedford,UK www.cranfield.ac.uk

Joint International Laser Conference 21 - 23 September, Edinburgh, UK http://clansman.com/lasers2003/

13th Annual Congress of the European Respiratory Society 27 September - 1 October, Vienna, Austria www.ersnet.org

6th Annual International Conference on Medical Image Computing and Computer-assisted Intervention 1 - 4 November, Toronto, Canada www.miccai2003.org

Annual Scientific Sessions of the American Heart Association 9 - 12 November, Orlando, FL, USA www.americanheart.org

35th British Medical Ultrasound Society Annual Scientific Meeting and Exhibition 10 - 12 December, Harrogate, UK http://www.bmus.org/

2004

IEE Symposium on Appropriate Medical Technology for Developing Countries 4 February, London, UK http://www.iee.org/Events/g04feb04.cfm

Eighth International Symposium on 3-D Analysis of Human Movement 31 March - 2 April, Tampa, FL, USA http://pe.usf.edu/isb3d/

Euroson 2004 -- 16th Congress of the European Federation of Societies for Ultrasound in Medicine and Biology 5 - 8 June, Zagreb, Croatia http://www.efsumb.org/

BIOSIGNAL 2004 -- 17th International EURASIP Conference 23 - 25 June, Brno, Czech Republic http://wes.feec.vutbr.cz/UBMI/bs2004.html

41st Annual Conference of the American Society for Healthcare Engineering 26 - 28 July, Orlando, FL, USA www.hospitalconnect.com/ashe/conference/welcome.html

26th Annual International Conference of the IEEE Engineering in Medicine and Biology Society 1 - 4 September, San Francisco, CA, USA www.eng.unsw.edu.au/embs/confs.html

Annual Congress of the European Society of Cardiology 28 August - 1 September, Munich, Germany www.escardio.org

36th British Medical Ultrasound Society Annual Scientific Meeting and Exhibition 8 - 10 December, Manchester, UK http://www.bmus.org/

2005

Euroson 2005 26 - 29 June, 17th Congress of the European Federation of Societies for Ultrasound in Medicine and Biology, Geneva, Switzerland http://www.efsumb.org/

42nd Annual Conference of The American Society for Healthcare Engineering 11 - 13 July, Anaheim, CA, USA www.hospitalconnect.com/ashe/conference/welcome.html

2006

Euroson 2006 3 - 6 October, 18th Congress of the European Federation of Societies for Ultrasound in Medicine and Biology, Bologna, Italy http://www.efsumb.org/