Electrical Impedance Tomography (EIT) and Its Medical Applications - A Review PDF
Electrical Impedance Tomography (EIT) and Its Medical Applications - A Review PDF
Electrical Impedance Tomography (EIT) and Its Medical Applications - A Review PDF
Abstract - This paper reviews the principles of Electrical Measuring local internal temperature changes associated
Impedance Tomography (EIT), different types of current patterns with hyperthermia treatments or cryosurgery, and improving
and reconstruction algorithms to assess its potential in medical electrocardiograms and electroencephalograms.
imaging. A current injection pattern in EIT has its own current In medical applications, due to the differences in
distribution profile within the subject under test. Hence, different
bioelectrical properties between tissues, the impedance
current patterns have different sensitivity, spatial resolution and
distinguishability. Image reconstruction studies with subject or distribution can indicates the structural and functional
practical phantoms are essential to assess the performance of properties of the object. In EIT, the current flow is
EIT systems for their validation, calibration and comparison determined by the impedance distribution within the object.
purposes. Impedance imaging of real objects or tissue phantoms As well, the problem is ill-posed which meaning that large
with different current injection methods is also essential for changes in impedance at the interior of the object can result
better assessment of the biomedical EIT systems. More in only small voltage changes at the surface.
specifically, this work reviews the three different image EIT equipment has relatively low cost and good portability
reconstruction techniques including back-projection method, and it is easy to be operated and maintained. The current
iterative method and one-step linearized method. In this review,
stimulation is not hazardous to humans in contrast to
Resistivity images are reconstructed from the boundary data
using Electrical Impedance Tomography and Diffuse Optical exposure to x-ray or radioisotopes in nuclear medicine. EIT
Tomography Reconstruction Software (EIDORS). shows potential to be used as a bedside real-time monitoring
system which is affordable in hospitals and clinics.
KEYWORDS: Electrical Impedance Tomography (EIT),
Stimulation patterns, regularization, Images. II. METHODS AND MATERIALS
I. INTRODUCTION
Electrical impedance tomography (EIT) [1] is a non-
invasive imaging modality that can be used to image
conductive subjects. In EIT the internal conductivity
distribution of the subject is reconstructed based on
electrical measurements from electrodes attached around the
boundary. In EIT, electrodes are attached on the surface of a
subject and a certain current pattern is injected into the
subject through stimulation electrodes. Commonly
alternating current is used as stimulation whose amplitude is Figure 1. Block diagram of a very simple EIT system which
usually several mA with frequency between 1~100 kHz. The shows main components.
voltages are measured using voltage measurement
electrodes. An image reconstruction method is then used to An EIT system mainly consists of two parts, which is data
calculate the internal conductivity distribution from the acquisition hardware and image reconstruction software.
boundary data. Figure 1 shows the basic block diagram of a very simple
EIT has numerous practical applications, e.g. in quality EIT system which comprised of
control and fault detection for various materials [2], 1. An array of electrodes, attached to the surface of the
geological exploration [3] and medical imaging [4]. In object
present days, EIT has been extensively researched in 2. A current generator is used for generating a constant
medical applications [5-9]. The possible medical sinusoidal alternating current to inject into the surface
applications for EIT [10] are monitoring for lung problems, 3. To perform data acquisition, a multiplexer circuit is
such as accumulating fluid or a collapsed lung [11], necessary for switching the current injector and
noninvasive monitoring of heart function and blood flow, voltmeter among the different data channels.
monitoring for internal bleeding, screening for breast cancer, 4. Demodulator/voltmeter is another important part of the
studying emptying of the stomach, studying pelvic fluid system for voltage measurement from the surface.
accumulation as a possible cause of pelvic pain, quantifying 5. For cancelling the noise, filters are used and ADC is
severity of premenstrual syndrome by determine the amount used for converting to digital signal and interfacing to
of intracellular versus extracellular fluid, determining the the Computer.
boundary between dead and living tissue. 6. An Image Reconstruction algorithm, which reconstructs
images of internal resistivity distribution of the object
Manuscript received September, 2013.
from the voltage measurements.
Dr.R.Harikumar, Professor, ECE Department, Bannari Amman
Institute of Technology, Sathyamangalam India. 7. A computer (display, hard copy, keyboard, storage, etc.)
R.Prbau, Assistant Professor, Biomedical Engineering Department, is used to enable an easy user access to the image
Velalar College of Engineering and Technology, Erode, India. analysis.
Dr.S.Raghavan, Professor, ECE Department, National Institute of
Technology, Tiruchirappalli, India.
193
Electrical Impedance Tomography (EIT) and Its Medical Applications: A Review
A. Stimulation And Measurement between the injecting electrodes, and decreases rapidly as a
Electrical impedance tomography reconstructs the function of distance. The method is as a result very sensitive
conductivity and permittivity within an object based on the to conductivity contrasts near the boundary and insensitive
conditions of voltage and current on the surface of the to central contrasts. It is also sensitive to perturbations in the
object. To find out the conditions of the current and voltage boundary shape of the object, in the positioning of the
on the surface, we can inject the current through a set of the electrodes and is quite sensitive to measurement error and
electrodes in contact with the surface and measure the noise [1].
developed voltages. There are many ways in the current- C. Opposite Method
injecting electrodes, voltage measuring electrodes and
The opposite or polar drive pattern [13], which is commonly
current patterns. This section describes various methods by
used in brain EIT [20], applies current through electrodes
which the current is injected and the voltages are measured.
that are 180◦ apart while voltage differences are measured on
B. Adjacent Drive Method the remaining electrodes. Hence this method is known as
The adjacent drive method [12], also called as the Opposite Method. Voltage differences are measured on the
neighbouring method, is the most common current driven voltage electrodes with respect to the electrode (called as the
pattern. In this method, the current is applied through voltage reference electrode) adjacent to the current-injecting
adjacent electrodes and the voltage is measured sequentially electrode (Figure.3)
from all other adjacent electrode pairs without the pairs
containing one or both the current electrodes. Figure.2
shows the adjacent method for a 16-electrode EIT system
with a circular domain surrounded by 16 surface electrodes
named as the electrode-1 to electrode-16. In adjacent
method, the first current projection is P1, the current is
injected through electrode-1 (1) and electrode-2 (2) and the
voltages differences (V1,V2,V3, . . ., V13) are measured
sequentially with 13 electrode pairs 3–4, 4–5, . . . and 15–
16. Voltages are not measured between pairs (16-1), (1-2),
or (2-3) (Figure.2.2a). Therefore the first current projection
(P1) gives 13 differential voltage data. Figure 3. Opposite current Driven pattern (a) first current
projection (P1) (b) second current projection (P2)
194
International Journal of Soft Computing and Engineering (IJSCE)
ISSN: 2231-2307, Volume-3, Issue-4, September 2013
against the other 13 electrodes. Hence, the current is injected fundamental problem in the image reconstruction is that, the
through electrodes- 16 (16) and electrodes-2 (2) and the electric current cannot be forced to flow linearly in an
differential voltages (V1, V2, V3,.., V13) are measured inhomogeneous volume conductor. Since there are no
sequentially with 13 electrode pairs 1–3, 1–4,.. and 1–15 sources within the volume conductor (the sources all lie on
(Fig.4a) considering electrode-1 (1) as the voltage reference the applied currents) then the potential field. The
(Fig.4a). Therefore the cross method gives 13 differential mathematical model describing the electrical properties of
voltage data. Next current is applied to electrodes 16 and 4 the field is a generalized Lap lace’s equation:
while 13 voltage measurements are taken using electrode 1
as the reference. This is repeated for currents injected (1)
between electrodes (16-4), (16-8), (16-10), (16-12), (16-14).
The entire procedure yields 7 X 13 = 91 measurements.
where ρ is the resistivity distribution; ϕ is voltage and D is
the tested field. Equation (1) indicates that there are no
current sources in the body.
(2)
(3)
195
Electrical Impedance Tomography (EIT) and Its Medical Applications: A Review
This method requests a full reconstruction of the the expected value of image. are modelled a
conductivity distribution using iterative approach which
priori.
considers intrinsic nonlinearity. The linearization of EIT
forward model is valid only under the condition that the Let . Here is the
conductivity changes are of low amplitudes. This condition mean of noise amplitude and is the a priori amplitude of
holds for small changes below 20% of the reference value. image element values. The measurement accuracy is
However, for some physiological activities such as modelled by W. For uncorrelated noise, each diagonal
pulmonary ventilation during deep breathing, this condition element of W is proportional to the corresponding channel
is not valid any more. The nonlinear method is applicable signal-to-noise-ratio (SNR). This one-step linearized
for these cases (with certain conditions apply). reconstruction gives results in a fast solution which is
applicable for real time functional imaging. But, the
It works as follows (as shown in Figure 5):
disadvantage is less accurate compare with iterative method.
First an estimation of the conductivity distribution;
Calculate the estimated voltage values through forward III. DISCUSSION.
solution;
Compare the estimated values with the original The images are reconstructed for this review from the
recorded voltage data and the error is used to calculate boundary data collected from the phantom or object with
the deviation between the real and the estimated different current patterns using Electrical Impedance and
conductivities; Diffuse Optical reconstruction Software (EIDORS) [20-21].
The conductivity vector estimated is then be adjusted. EIDORS is open source software which is MatLab based
This procedure is repeated until the error between the image reconstruction algorithm. EIDORS is used to
estimated and recorded voltages is minimized to an reconstruct the images from electrical or diffuse optical data
acceptable level. which is developed with Gauss Newton method. Forward
problem and inverse problems are solved with a Finite
element method (FEM). For this review, the EIT data’s are
taken from EIDORS for creating the reconstruction image.
(4)
It addresses the inverse solution as a linear reconstruction
matrix B and allows use of advanced regularization methods
to solve the inverse problem. The Gauss-Newton (GN)
method in EIT [17-19] estimates a solution by minimizing
(5)
Figure.7. Voltage pattern for opposite stimulation
Where is the covariance matrix of the
The voltage pattern for adjacent stimulation which shows
measurement noise n. Since noise channels are independent, (Figure.6.) that the current is injected in the adjacent
Σn is a diagonal matrix with , where σi2 is the electrodes and potentials are measured in remaining adjacent
electrodes. In this pattern the forward problem is solved
noise variance at channel. Here is the
using FEM models. Figure.7. shows that the voltage pattern
covariance matrix of the desired image and x0 represents for opposite stimulation, in which the current is injected in
196
International Journal of Soft Computing and Engineering (IJSCE)
ISSN: 2231-2307, Volume-3, Issue-4, September 2013
the opposite electrodes and potential differences are available in market. . But In the application of control
measured on the voltage electrodes with respect to the ventilation, an automatic system could be useful in order to
reference electrode adjacent to the current-injecting better manage of Acute and Weaning phases of mechanical
electrode. ventilation. To enabling the clinically useful automatic
In case of thoracic imaging, EIT is used to measuring the system,
ventilation, perfusion and gas exchange. EIT images reflect The systems should be readily available at a affordable
the lung function which means EIT displays ventilated lung cost
regions rather than morphological or anatomical structures EIT data and images should be accessible in standard
of the lung. EIT images are contained regional information, formats
when pathological conditions like pleural effusion or EIT should be robust against electrode contact problems
atelectasis lead to non-aerated and non-ventilated lung and electrical interference problems.
regions. Meanwhile EIT images display lungs regions with Reconstruction Software should have a spontaneous
trapped air (e.g. pneumothorax) in black because they are interface focused on the real-time clinical user.
not ventilated and also partially or not ventilated regions are To enhance clinical decision making, automated
displayed in dark blue or black color. Figure 8. shows the approaches for data analysis and interpretation must be
some of the examples of EIT status images [23]. included.
In the future, EIT is almost undoubtedly useful in the
diagnosis of pulmonary problems, especially for heavily
instrumented patients in the ICU to whom the compactness
and bedside suitability of EIT are most helpful.
REFERENCES
1. J.G. Webster, Electrical Impedance Tomography. Adam Hilger Series
of Biomedical Engineering, Adam Hilger, New York, USA, 1990.
2. S. Feliu, J.C. Galvan, and M. Morcillo, "interpretation of electrical
impedance diagram for painted galvanized steel," Prog. in Organic
Coatings, vol. 17, no. 2, pp.143- 153, 1989.
3. K.A.Dines and R.J.Lytle, "Analysis of ele-ctrical conductivity
imaging," Geophysics, vol. 46, no. 7, pp. 1025- 1036, July 1981.
4. D.C.Barber, B.H. Brown, and J. L. Freeston, "Imaging spatial
distributions of resistivity using applied potential tomography,"
Electron. Lett., vol. 19, pp. 933 - 935, 1983.
5. B.H.Brown, “Medical impedance tomography and process impedance
tomography: a brief review”, Measure. Sci. Technol. 12 (August)
(2001) 991–996.
6. A.P.Bagshaw, A.D.Liston, R.H.Bayford, A. Tizzard, A.P.Gibson,
A.T.Tidswell, M.K. Sparkes, H.Dehghani, C.D.Binnie, D.S. Holder,
“Electrical impedance tomography of human brain function using
reconstruction algorithms based on the finite element method”,
NeuroImage 20 (2003) 752–764.
7. F.S. Moura, J.C.C. Aya, A.T. Fleury, M.B.P. Amato, R.G. Lima,
“Dynamic imaging in electrical impedance tomography of the human
chest with online transition matrix identification”, IEEE Trans.
Biomed. Eng. 57 (2) (2010).
8. F. Ferraioli, A. Formisano, R. Martone, “Effective exploitation of
prior information in electrical impedance tomography for thermal
monitoring of hyperthermia treatments”, IEEE Trans. Mag. 45 (3)
(2009).
9. A.E.Hoetink, T.J.C.Faes, J.T. Marcus, H.J.J. Kerkkamp, R.M.
Heethaar, “Imaging of thoracic blood volume changes during the
Figure 8. EIT images with various status of lung conditions. heart cycle with electrical impedance using a linear spot-electrode
array”, IEEE Trans. Med. Imaging 21 (6) (2002) 653.
From the left to right and top to bottom, the tidal image of 10. D. Holder, “Clinical and Physiological Applications of Electrical
a) healthy individual b) patient with pleural effusion in the Impedance Tomography”. London, U.K.: UCL Press, 1993.
left lung c) patient with pneumectomy d) patient with a 11. J.C. Newell, D. Isaacson, G.J. Saulnier, M. Cheney, and D.G. Gisser,
pneumothorax in the left lung e) patient with dorsal “Acute pulmonary edema assessed by electrical impedance
tomography,” in Proc. Annui. Int. Conf. IEEE Engineering in
atelectasis f) ARDS patient Medicine and Biology Soc., 1993, pp. 92-93.
12. B.H. Brown, A.D. Segar, “The Sheffield data collection system”, Clin
IV. CONCLUSION Phys Physiol Measur 8 (Suppl. A) (1987) 91–97.
13. Avis NJ, Barber DC, “Image reconstruction using non-adjacent drive
EIT system is having a great potential to provide a new configuration”, Physiol Meas, 1994, 16, A153-A160.
medical imaging modality which is radiation free, non- 14. Hua P, Webster JG, Tompkins WJ,” Effect of the measurement
method on noise handling and image quality of EIT imaging”, IEEE
nephrotoxic, portable and inexpensive. Existing EIT was 9th Annual Conf on Engineering in Medicine and Biological Science,
useful to assessment of regional lung aeration and tidal 1993, 1429-30.
volume and monitoring rapid lung volume changes. 15. Gisser G, Isaacson D, Newell JC, “Current topics in impedance
Currently EIT systems are capable of making measurement imaging”, Clin Physiol Meas, 1987, 8:38-46.
16. D.C.Barber and B.H.Brown “Recent developments in applied
at much higher rates and recent image reconstruction potential tomography”-apt Information Processing in Medical
algorithms provide much improved resolution and Imaging 1986 pp 106–21
robustness to electrode errors with reduced image artifacts 17. T.J.Yorkey, J.G.Webster, and W.J. Tompkins. “Comparing
and also have greatly reduced noise and interference reconstruction algorithms for electrical impedance tomography”.
IEEE Trans. Biomed. Eng, 34:843–852, 1987.
systems. Some of the commercial systems are currently
197
Electrical Impedance Tomography (EIT) and Its Medical Applications: A Review
198