Use of statistical parametric mapping (SPM) to enhance electrical impedance tomography (EIT) image sets

Use of statistical parametric mapping (SPM), which is widely used in analysis of neuroimaging studies with fMRI and PET, has the potential to improve quality of EIT images for clinical use. Minimal modification to SPM is needed, but statistical analysis based on height, not extent thresholds, should be employed, due to the 20-80% variation of the point spread function, across EIT images. SPM was assessed in EIT images reconstructed with a linear time difference algorithm utilizing an anatomically realistic finite element model of the human head. Images of the average of data sets were compared with those produced using SPM over 10-40 individual image data sets without averaging. For a point disturbance, a sponge 15% of the diameter of an anatomically realistic saline-filled tank including a skull, with a contrast of 15%, and for visual evoked response data in 14 normal human volunteers, images produced with SPM were less noisy than the average images. For the human data, no consistent physiologically realistic changes were seen with either SPM or direct reconstruction; however, only a small data set was available, limiting the power of the SPM analysis. SPM may be used on EIT images and has the potential to extract improved images from clinical data series with a low signal-to-noise ratio.

Improving the finite element forward model of the human head by warping using elastic deformation

As the use of realistic geometry in the forward model of electrical impedance tomography (EIT) of brain function appears to improve image reconstruction, the generation of patient-specific finite element meshes has been the subject of much recent work. This paper presents a more rapid method of generating more geometrically accurate finite element meshes of the human head by warping existing meshes such that the surface boundary beneath the electrodes closely matches that of the subject with minimal degradation to the quality of the mesh. Pre-existing meshes of spheres and adult head models incorporating key internal anatomical features are warped, using elastic deformation, to match a phantom latex tank incorporating a real skull. The algorithm is described and tests are carried out to optimize the key parameters to ensure minimal degradation of mesh quality and distortion of internal features. Results show that the algorithm operating with the optimum parameters produces meshes of sound quality and could represent an important step in the timely and productive creation of forward models in clinical applications.

Bioimpedance tomography (electrical impedance tomography)

Electrical impedance tomography (EIT) is a relatively new imaging method that has evolved over the past 20 years. It has the potential to be of great value in clinical diagnosis; however, EIT is a technically difficult problem to solve in terms of developing hardware for data capture and the algorithms to reconstruct the images. This review looks at the development of EIT and how it has evolved. It focuses on its clinical applications, examining hardware for the collection of data and reconstruction algorithms to generate images. Finally, this review looks at future developments that are evolving from EIT. These new variations use mixed modalities that may produce interesting new clinical imaging tools.

Factors limiting the application of electrical impedance tomography for identification of regional conductivity changes using scalp electrodes during epileptic seizures in humans

Electrical impedance tomography (EIT) has the potential to produce images during epileptic seizures. This might improve the accuracy of the localization of epileptic foci in patients undergoing presurgical assessment for curative neurosurgery. It has already been shown that impedance increases by up to 22% during induced epileptic seizures in animal models, using cortical or implanted electrodes in controlled experiments. The purpose of this study was to determine if reproducible raw impedance changes and EIT images could be collected during epileptic seizures in patients who were undergoing observation with video-electroencephalography (EEG) telemetry as part of evaluation prior to neurosurgery to resect the region of brain causing the epilepsy. A secondary purpose was to develop an objective method for processing and evaluating data, as seizures arose at unpredictable times from a noisy baseline. Four-terminal impedance measurements from 258 combinations were collected continuously using 32 EEG scalp electrodes in 22 seizure episodes from 7 patients during their presurgical assessment together with the standard EEG recordings. A reliable method for defining the pre-seizure baseline and recording impedance data and EIT images was developed, in which EIT and EEG could be acquired simultaneously after filtering of EIT artefact from the EEG signal. Fluctuations of several per cent over minutes were observed in the baseline between seizures. During seizures, boundary voltage changes diverged with a standard deviation of 1-54% from the baseline. No reproducible changes with the expected time course of some tens of seconds and magnitude of about 0.1% could be reliably measured. This demonstrates that it is feasible to acquire EIT images in parallel with standard EEG during presurgical assessment but, unfortunately, expected EIT changes on the scalp of about 0.1% are swamped by much larger movement and systematic artefact. Nevertheless, EIT has the unique potential to provide invaluable neuroimaging data for this purpose and may still become possible with improvements in electrode design and instrumentation.

Multi-frequency electrical impedance tomography (EIT) of the adult human head: initial findings in brain tumours, arteriovenous malformations and chronic stroke, development of an analysis method and calibration

MFEIT (multi-frequency electrical impedance tomography) could distinguish between ischaemic and haemorrhagic stroke and permit the urgent use of thrombolytic drugs in patients with ischaemic stroke. The purpose of this study was to characterize the UCLH Mk 2 MFEIT system, designed for this purpose, with 32 electrodes and a multiplexed 2 kHz to 1.6 MHz single impedance measuring circuit. Data were collected in seven subjects with brain tumours, arteriovenous malformations or chronic stroke, as these resembled the changes in haemorrhagic or ischaemic stroke. Calibration studies indicated that the reliable bandwidth was only 16-64 kHz because of front-end components placed to permit simultaneous EEG recording. In raw in-phase component data, the SD of 16-64 kHz data for one electrode combination across subjects was 2.45 +/- 0.9%, compared to a largest predicted change of 0.35% estimated using the FEM of the head. Using newly developed methods of examining the most sensitive channels from the FEM, and nonlinear imaging constrained to the known site of the lesion, no reproducible changes between pathologies were observed. This study has identified a specification for accuracy in EITS in acute stroke, identified the size of variability in relation to this in human recordings, and presents new methods for analysis of data. Although no reproducible changes were identified, we hope this will provide a foundation for future studies in this demanding but potentially powerful novel application.

Using the GRID to improve the computation speed of electrical impedance tomography (EIT) reconstruction algorithms

In our group at University College London, we have been developing electrical impedance tomography (EIT) of brain function. We have attempted to improve image quality by the use of realistic anatomical meshes and, more recently, non-linear reconstruction methods. Reconstruction with linear methods, with pre-processing, may take up to a few minutes per image for even detailed meshes. However, iterative non-linear reconstruction methods require much more computational resources, and reconstruction with detailed meshes was taking far too long for clinical use. We present a solution to this timing bottleneck, using the resources of the GRID, the development of coordinated computing resources over the internet that are not subject to centralized control using standard, open, general-purpose protocols and are transparent to the user. Optimization was performed by splitting reconstruction of image series into individual jobs of one image each; no parallelization was attempted. Using the GRID middleware 'Condor' and a cluster of 920 nodes, reconstruction of EIT images of the human head with a non-linear algorithm was speeded up by 25-40 times compared to serial processing of each image. This distributed method is of direct practical value in applications such as EIT of epileptic seizures where hundreds of images are collected over the few minutes of a seizure and will be of value to clinical data collection with similar requirements. In the future, the same resources could be employed for the more ambitious task of parallelized code.

Generating accurate finite element meshes for the forward model of the human head in EIT

The use of realistic anatomy in the model used for image reconstruction in EIT of brain function appears to confer significant improvements compared to geometric shapes such as a sphere. Accurate model geometry may be achieved by numerical models based on magnetic resonance images (MRIs) of the head, and this group has elected to use finite element meshing (FEM) as it enables detailed internal anatomy to be modelled and has the capability to incorporate information about tissue anisotropy. In this paper a method for generating accurate FEMs of the human head is presented where MRI images are manually segmented using custom adaptation of industry standard commercial design software packages. This is illustrated with example surface models and meshes from adult epilepsy patients, a neonatal baby and a phantom latex tank incorporating a real skull. Mesh quality is assessed in terms of element stretch and hence distortion.

The effect of layers in imaging brain function using electrical impedance tomograghy

Electrical impedance tomography (EIT) has promise for imaging brain function with rings of scalp electrodes, but hitherto human images have been collected and reconstructed using a simple algorithm in which the head was modelled as a homogeneous sphere. The purpose of this work was to assess the improvement in image quality which could be achieved by adding layers to represent the cerebro-spinal fluid (CSF), skull and scalp in the forward model employed by the reconstruction algorithm. Solutions to the forward model were produced analytically and using the linear finite element method (FEM). This was undertaken for computer simulated data when a spherical conductivity change of 10%, radius 5 mm, was moved through 29 positions within a head modelled as four concentric spheres of radius 80-92 mm in order to verify the accuracy of the linear FEM by comparison with the analytical method. Test data were also recorded in a 93.5 mm, spherical, saline-filled tank in which the skull was simulated by a hollow sphere of plaster of Paris, 5 mm thick and a 20 x 20 mm right-cylindrical Perspex object, a 100% conductivity decrease, was moved through 39 positions. The best images were achieved by reconstruction with a four- or three-shell analytical model, giving a spatial accuracy of 5.8 +/- 2.2 mm for computer simulated or 14.0 +/- 5.8 mm for tank data. Mean FWHM was 57 mm and 91 mm in the XY-plane and along the z-axis, respectively. Reconstruction with a homogeneous analytical model gave localization errors greater by about 50-300%, but a reduction in FWHM of about 5% of the image diameter. Unexpectedly, reconstruction with FEM models gave poorer results similar to the analytical homogeneous case. This confirms that addition of shells to the forward model improves image quality as expected with an analytical model for reconstruction, but that the FEM method employed, which used a medium mesh and a linear element computation, requires improvement in order to yield the expected benefits.

Neuromagnetic field strength outside the human head due to impedance changes from neuronal depolarization

The holy grail of neuroimaging would be to have an imaging system, which could image neuronal electrical activity over milliseconds. One way to do this would be by imaging the impedance changes associated with ion channels opening in neuronal membranes in the brain during activity. In principle, we could measure this change by using electrical impedance tomography (EIT) but it is close to its threshold of detectability. With the inherent limitation in the use of electrodes, we propose a new scheme based on recording the magnetic field resulting from an injected current with superconducting quantum interference devices (SQUIDs), used in magnetoencephalography (MEG). We have performed a feasibility study using computer simulation. The head was modelled as concentric spheres to mimic the scalp, skull, cerebrospinal fluid and brain using the finite element method. The magnetic field 1 cm away from the scalp was estimated. An impedance change of 1% in a 2 cm radius volume in the brain was modelled as the region of depolarization. A constant current of 100 microA was injected into the head from diametrically opposite electrodes. The model predicts that the standing magnetic field is about 10 pT and changed by about 3 fT (0.03%) on depolarization. The spectral noise density in a typical MEG system in the frequency band 1-100 Hz is about 7 fT, so this places the change at the limit of detectability. This is similar to electrical recording, as in conventional EIT systems, but there may be advantages to MEG in that the magnetic field directly traverses the skull and instrumentation errors from the electrode-skin interface will be obviated.

Electrical impedance tomography spectroscopy (EITS) for human head imaging

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 comparison of headnet electrode arrays for electrical impedance tomography of the human head

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.

A comparison of techniques to optimize measurement of voltage changes in electrical impedance tomography by minimizing phase shift errors

In electrical impedance tomography, errors due to stray capacitance may be reduced by optimization of the reference phase of the demodulator. Two possible methods, maximization of the demodulator output and minimization of reciprocity error have been assessed, applied to each electrode combination individually, or to all combinations as a whole. Using an EIT system with a single impedance measuring circuit and multiplexer to address the 16 electrodes, the methods were tested on resistor-capacitor networks, saline-filled tanks and humans during variation of the saline concentration of a constant fluid volume in the stomach. Optimization of each channel individually gave less error, particularly on humans, and maximization of the output of the demodulator was more robust. This method is, therefore, recommended to optimize systems and reduce systematic errors with similar EIT systems.

Design and performance of the UCLH mark 1b 64 channel electrical impedance tomography (EIT) system, optimized for imaging brain function

The UCLH Mark 1b is a portable EIT system that can address up to 64 electrodes, which has been designed for imaging brain function with scalp electrodes. It employs a single impedance-measuring circuit and multiplexer so that electrode combinations may be addressed flexibly using software. It operates in the relatively low frequency band between 225 Hz and 77 kHz, as lower frequencies produce larger changes during brain activity, and has a videocassette-sized headbox on a lead 10 m long, connected to a base box the size of a video recorder, and notebook PC, so that recordings may be made in ambulant subjects. Its performance was assessed using a resistor-capacitor network, and two saline-filled tanks-a cylindrical Perspex one and a latex one which contained a human skull. System signal-to-noise ratio was better than 50 dB and the maximum reciprocity error less than 10% for most frequencies. The CMMR was better than 80 dB at 38 kHz and a sponge, 20 mm across, which caused a local 12% impedance increase, was correctly localized in images. This suggests that the system has adequate performance to image impedance changes of 5-50% known to occur in the brain during normal activity, epilepsy or stroke; clinical trials to image these conditions are in progress.

A multi-shell algorithm to reconstruct EIT images of brain function

Electrical impedance tomography (EIT) may be used to image brain function, but an important consideration is the effect of the highly resistive skull and other extracerebral layers on the flow of injected current. We describe a new reconstruction algorithm, based on a forward solution which models the head as four concentric, spherical shells, with conductivities of the brain, cerebrospinal fluid, skull and scalp. The model predicted that the mean current travelling in the brain in the diametric plane for current injection from polar electrodes was 5.6 times less than if the head was modelled as a homogeneous sphere; this suggests that an algorithm based on this should be more accurate than one based on a homogeneous sphere model. In images reconstructed from computer-simulated data or data from a realistic saline-filled tank containing a real skull, a Perspex rod was localized to within 17% or 20% of the tank diameter of its true position, respectively. Contrary to expectation, the tank images were less accurate than those obtained with a reconstruction algorithm based on a homogeneous sphere. It is not yet clear if the theoretical advantages of this algorithm will yield practical advantages for head EIT imaging; it may be necessary to proceed to more complex algorithms based on numerical models which incorporate realistic head geometry. If so, this analytical forward model and algorithm may be used to validate numerical solutions.

Validation of a 3D reconstruction algorithm for EIT of human brain function in a realistic head-shaped tank

Previous work has demonstrated that electrical impedance tomography can be used to image human brain activity during evoked responses, but two-thirds of the reconstructed images fail to localize an impedance change to the expected stimulated cortical area. The localization failure may be caused by modelling the head as a homogenous sphere in the reconstruction algorithm. This assumption may lead to errors when used to reconstruct data obtained from the human head. In this study a 3D reconstruction algorithm, based on a model of the head as a homogenous sphere, was characterized by simulating the algorithm model, the head shape and the presence of the skull in saline-filled tanks. EIT images of a sponge, 14 cm3 volume with a resistivity contrast of 12%, were acquired in three different positions in tanks filled with 0.2% saline. In a hemispherical tank, 19 cm in diameter, the sponge was localized to within 3.4-10.7% of the tank diameter. In a head-shaped tank, the errors were between 3.1 and 13.3% without a skull and between 10.3 and 18.7% when a real human skull was present. A significant increase in localization error therefore occurs if an algorithm based on a homogeneous sphere is used on data acquired from a head-shaped tank. The increased error is due to the presence of the skull, as no significant increase in error occurred if a head-shaped tank was used without the skull present, compared to the localization error within the hemispherical tank. The error due to the skull significantly shifted the impedance change within the skull towards the centre of the image. Although the increased localization error due to the skull is not sufficient to explain the localization errors of up to 50% of the image diameter present in the images of some human subjects, the future use of a realistic head model in the reconstruction algorithm is likely to reduce the localization error in the human images due to the presence of the skull.

Solving the forward problem in electrical impedance tomography for the human head using IDEAS (integrated design engineering analysis software), a finite element modelling tool

If electrical impedance tomography is to be used as a clinical tool, the image reconstruction algorithms must yield accurate images of impedance changes. One of the keys to producing an accurate reconstructed image is the inclusion of prior information regarding the physical geometry of the object. To achieve this, many researchers have created tools for solving the forward problem by means of finite element methods (FEMs). These tools are limited, allowing only a set number of meshes to be produced from the geometric information of the object. There is a clear need for geometrical accurate FEM models to improve the quality of the reconstructed images. We present a commercial tool called IDEAS, which can be used to create FEM meshes for these models. The application of this tool is demonstrated by using segmented data from the human head to model impedance changes inside the head.

Three-dimensional electrical impedance tomography of human brain activity

Regional cerebral blood flow and blood volume changes that occur during human brain activity will change the local impedance of that cortical area, as blood has a lower impedance than that of brain. Theoretically, such impedance changes could be measured from scalp electrodes and reconstructed into images of the internal impedance of the head. Electrical Impedance Tomography (EIT) is a newly developed technique by which impedance measurements from the surface of an object are reconstructed into impedance images. It is fast, portable, inexpensive, and noninvasive, but has a relatively low spatial resolution. EIT images were recorded with scalp electrodes and an EIT system, specially optimized for recording brain function, in 39 adult human subjects during visual, somatosensory, or motor activity. Reproducible impedance changes of about 0.5% occurred in 51/52 recordings, which lasted from 6 s after the stimulus onset to 41 s after stimulus cessation. When these changes were reconstructed into impedance images, using a novel 3-D reconstruction algorithm, 19 data sets demonstrated significant impedance changes in the appropriate cortical region. This demonstrates, for the first time, that significant impedance changes, which could form the basis for a novel neuroimaging technology, may be recorded in human subjects with scalp electrodes. The final images contained spatial noise and strategies to reduce this in future work are presented.

Electrical impedance tomography of human brain activity with a two-dimensional ring of scalp electrodes

Previously, electrical impedance tomography (EIT) has been used to image impedance decreases in the exposed cortex of rabbits during brain activity. These are due to increased blood volume at the site of the stimulated cortex; as blood has a lower impedance than brain, the impedance decreases. During human brain activity similar blood flow changes have been detected using positron emission tomography (PET) and functional magnetic resonance imaging (fMRI). If blood volume also changes then the impedance of human cortex will change during brain activity; this could theoretically be imaged with EIT. EIT data were recorded from a ring of 16 scalp electrodes in 34 recordings in 19 adult volunteers before, during and after stimulation with (1) a visual stimulus produced by an 8 Hz oscillating checkerboard pattern or (2) sensory stimulation of the median nerve at the wrist by a 3 Hz electrical square wave stimulus. Reproducible impedance changes, with a similar timecourse to the stimulus, were seen in all experiments. Significant impedance changes were seen in 21 +/- 5% (n = 16, mean +/- SEM) and 19 +/- 3% (n = 18) of the electrode measurements for visual and somatosensory paradigms respectively. The reconstructed 2D EIT images showed reproducible impedance changes in the approximate region of the stimulated cortex in 7/16 visual and 5/18 somatosensory experiments. This demonstrates that reproducible impedance changes can be measured during human brain activity. The final images contained spatial noise; the reasons for this and strategies to reduce this in future are discussed.

Two-dimensional finite element modelling of the neonatal head

Electrical impedance tomography (EIT) could allow the early diagnosis of infant brain injury following birth asphyxia. The purpose of this work was to determine the effect of variations in skull, scalp or cerebrospinal fluid (CSF) resistivity, as these vary in clinical conditions and could degrade image quality. These factors were investigated using finite element models of the adult and neonatal head. The results suggest that there is a wide range over which the resistivity of the neonatal skull has little effect on the sensitivity to a central impedance change. The scalp and CSF appear to shunt current away from the brain; when their resistivity was decreased from normal values, this shunting effect increased and caused a decrease in sensitivity to a central resistance change. The resistivity of neonatal skull has not, to our knowledge, been directly measured and will anyway vary within and between individuals; this work suggests that EIT will be relatively insensitive to variations in neonatal skull impedance.

Assessment and calibration of a low-frequency system for electrical impedance tomography (EIT), optimized for use in imaging brain function in ambulant human subjects

An EIT system has been produced that has been optimized for imaging impedance changes with scalp electrodes during brain activity in ambulant subjects. It can record from 225 Hz to 65 kHz, has a small headbox on a lead 10 m long, and has software programmable electrode selection. In calibration experiments in a small cylindrical tank filled with potassium chloride solution and samples of cucumber, noise was less than 1% with averaging, and acceptable images were produced at frequencies down to 1800 Hz. This suggests that EIT can be performed at low frequencies, which are likely to give larger signals during brain activity. Future work will include trials in humans and improvement of the current source and isolation.

Improvement of the positional accuracy of EIT images of the head using a Lagrange multiplier reconstruction algorithm with diametric excitation

A novel algorithm for the reconstruction of dynamic images using diametric excitation has been developed. The algorithm is specifically designed to image impedance changes in the brain using boundary data obtained from scalp electrodes by incorporating a priori information. The a priori information is obtain by solving the forward problem using a finite-element model (FEM) which includes the discontinuity of the skull resistivity. The advantages with this new approach are that the sensitivity and accuracy of the location of the impedance changes are improved compared to methods based on adjacent excitation.