A New Design of Electrical Impedance Tomography Sensor System for Pulmonary Disease Diagnosis

Xiaoyuan LIU, Shihong YUE, Zeying WANG

Journal of Systems Science and Information ›› 2018, Vol. 6 ›› Issue (5) : 473-480.

PDF(578 KB)
PDF(578 KB)
Journal of Systems Science and Information ›› 2018, Vol. 6 ›› Issue (5) : 473-480. DOI: 10.21078/JSSI-2018-473-08
 

A New Design of Electrical Impedance Tomography Sensor System for Pulmonary Disease Diagnosis

Author information +
History +

Abstract

As an advanced process detection technology, electrical impedance tomography (EIT) has wide application prospects and advantages in medical imaging diagnosis. However, a series of issues need to be addressed before applying EIT for bedside monitoring. Medical diagnosis and bedside monitoring are dynamic measuring process, where the positions of measuring electrodes and the shape of the detected field are changing dynamical. Due to the inability to cope with the changeable electrode positions and various dynamic fields, existing EIT systems are mainly used for industrial detection in condition of static measurement and visualization. In this paper, we investigate the dynamic measurement and visualization of human breast in EIT field, describe the design of the measuring sensor system, and expound the measuring principle. The main component of the hardware system is a builtin servo electrical resistance tomography sensor with capacitive sliding rod, which can adapt to the crowd of different chest contour and the change of chest shape in the dynamic process of breathing. The corresponding measuring principle is extracting all real-time positions of measuring electrodes, then obtaining the dynamic boundary, finally dividing the detection field rapidly. Experimental results confirmed that the proposed system can obtain real-time location of boundary sensor and dynamically solve the problem of arbitrary-shape boundary measurement. The imaging results validate the availability of designed sensor system and the effectiveness of the corresponding measuring principle.

Key words

electrical impedance tomography / non-invasive / dynamic boundary / human breast / sensor

Cite this article

Download Citations
Xiaoyuan LIU , Shihong YUE , Zeying WANG. A New Design of Electrical Impedance Tomography Sensor System for Pulmonary Disease Diagnosis. Journal of Systems Science and Information, 2018, 6(5): 473-480 https://doi.org/10.21078/JSSI-2018-473-08

1 Introduction

Electrical impedance tomography (EIT) technology is an advanced process detection technology with the characteristics of invasive, safety, real-time, and economy. EIT has been widely used in industry[1, 2], as well has a very promising application prospect in the fields of organism physiological function research, disease diagnosis and bedside monitoring[3, 4]. Medical diagnosis has always been valued[5], and EIT technology is desired to make great progress, because the ideal EIT can reconstruct real-time dynamic images of the distribution of impedance changes of biological tissues under different frequency and biological organs physiological activities (such as breath, heart beat). Moreover, it can also compensate for the existing defects of medical imaging technology in the field of functional imaging to achieve the purpose of early diagnosis and treatment of diseases. EIT technology has the ability to solve the problem of effective real-time visualization monitoring[6, 7].
In order to achieve the above goals, the following key problems of EIT technology need to be solved[8]. Firstly, there is the problem of dynamic measurement. In the human respiratory process, the measuring electrodes have to be attached to the skin surface of the human chest. Thus, the locations of these electrodes are moving dynamically. Secondly, as the measuring electrodes move, the size and shape of the measuring field will change dynamically. However, in the existing industrial detection applications, the measuring electrodes are attached on the boundary. In order to be adapted to the traditional measuring electrodes, the existing method assumes that the boundary of the measuring field is invariable. At present, in the human chest two-dimensional measurement of EIT technology, all of the tests and measuring electrodes (usually 16 or 32) must artificially be fixed and adjusted on the surface of the human thorax to keep them in a cross-section. In general, these electrodes are separately fixed on the surface of the chest, orientate the position of each electrode through the geometric measurement or other means of distance sensors, and then connect the measured field through all electrode positions. In this way, the cross section of the chest can be visualized by EIT technology.
There are three problems with the above existing measurement of the human thorax[9, 10]. Firstly, when these measurements and adjustments are repeatedly and continuously, all electrodes cannot distribute evenly among the same cross-section. Secondly, the shape and the size of the measuring field determined by electrodes are dynamic in the human respiration process. Therefore, the existing measuring methods cannot adapt to the dynamic changes of the measurement points, and it is difficult to pick up real-time dynamic data. In addition, electrode adjustment is also easy to break the connection points between electrodes and conductors, scuff electrodes, and break off the electrodes from the skin and cause other issues[11-13]. These problems seriously affect the service life of the electrodes and the accuracy of the measured signal.

2 Design of the New Sensor System

This section illustrates the structure of the new EIT sensor system and the measuring principles.

2.1 Structure of the New EIT Sensor System

The new EIT system, which consists of a 16-electrode array of capacitance sensors, is fixed in a steel plane with an ellipsoidal measuring area, as shown in Figure 1.
Figure 1 Manufactured sensor system

Full size|PPT slide

Each electrode is equipped on a sliding capacitance rod and its end is connected to the surface of measuring human chest, and the other end is connected to the EIT measuring and controlling system. One sliding capacitance sensor, including a metallic cylinder, a hollow non-metallic rod, a metallic rod, and lubricate centric mediums, as shown in Figure 2.
Figure 2 The structure of the new sensor

Full size|PPT slide

A pair of metal cylinders and metal rod serves as fixed and movable capacitance electrodes to measure the capacitance values. The two ends of the metal rod are connected to the metal cylinder and the nonmetal hollow rod respectively. A signal wire is connected to the external measuring circuit through the hollow non-metallic rod. The press rid of the connecting lines is fixed on the metallic rod. In addition, the hollow metallic rod and the metallic rod are also used to adjust the length of the entire sliding rod. In order to ensure the accuracy of the permittivity, centric mediums apply the Teflon, whose friction coefficient is the smallest of all plastic materials, then the metal rod can be moved freely into the metallic cylinder. The press rod is used to fix the metallic cylinder and these centric media and then can be updated easily. All the above components are placed in the same plane. The sliding rods can freely slide along a sliding way. The movement of the sliding rod can generate different capacitance values corresponding to the position of the rod.

2.2 Measuring Principle of the New System

As illustrated in Figure 3, the initial capacitance of the concentric cylindrical capacitance displacement sensor C0 is given by
C0=2πεrL0/(lnD0lnD1).
(1)
Figure 3 Concentric cylindrical capacitance sensor

Full size|PPT slide

The capacitance changes when the entire metallic rod moves relative to the fixed metallic cylinder in the human respiratory process. Suppose the displacement of the metallic rods is x, and the generating capacitance value between the metallic rod and the metal cylinder Cx is given by
Cr=2πεr(L0x)/(lnD0lnD1).
(2)
With reference to Equation (1) and Equation (2)
Cx=C0(1xL0),
(3)
where εr is the product of the permittivity of vacuum and relative dielectric constant of the medium between metallic rod and metallic cylinder; x is the displacement of the metallic rods; L0 is the half of the metal cylinder length; D0 is the inner diameter of the metallic cylinder; D1 is the external diameter of the metallic cylinder.
During the breathing process, as the thoracic cage expanding and contracting, the sliding rods closely stick with the chest, and the position between the metal rods and the metallic cylinder changes relatively. With reference to Equation (3), the linear relationship of Cx and x can be obtained, the displacement of the metal rods and the capacitance between metallic rods and metallic cylinder. Hence, the change of displacement can be obtained by the change in capacitance.
The magnitude of the output voltage in the actual measurement process reflects the change of the measured capacitance. Measured capacitance and voltage converting circuit are shown in Figure 4, where Vi(t) represents excitation source, CS1 and CS2 represent the grounded capacitances distributed on both ends of the measured capacitance, Cx represents the capacitance under the testing process, Cf represents the feedback capacitance, Rf represents the feedback resistance, Vo(t) represents the output voltage of the operational amplifier.
Figure 4 C/V converting circuit

Full size|PPT slide

When |jωCfRf|1, the output voltage of C/V converting circuit is given by
V0=CxCfVi.
(4)
The C/V converting circuit is shown in Figure 4. The system generates a voltage signal proportional to the measured capacitance. As long as the output voltage amplitude is known, the value of measured capacitance can be derived. With reference to Equation (3) and Equation (4):
V0=C0Cf(1xL0)Vi.
(5)
C0 (the initial capacity of concentric cylindrical capacitance displacement sensor) and Cf (the feedback capacitance of the C/V converting circuit) are both invariable. Putting k=C0/Cf into Equation (5) results in
V0=k(1xL0)Vi.
(6)
It is obvious that the displacement and the output voltage have a linear relationship when excitation voltage is fixed. Then the real-time changes of any chest contour through the location information can be detected by the displacements of 16 concentric metallic rods.
In the human respiratory process, the electrodes attached to the human chest drive the sliding rod moving in the corresponding sliding approach, and then connect each sliding rod end in the ellipsoidal shell to determine the boundary of human chest at any time. In fact, the scale length of 32 sliding rods on an ellipsoidal shell can be solved by capacitance of each capacitor sliding rod. Ellipsoidal shell length of the long axis and short axis is known, and each sliding rod is ellipsoidal circle evenly distributed. Therefore, when the ellipsoidal boundary points are known in prior, the chest boundary can be determined by the relative position of each electrode, as shown in Figure 5. Then the locations of these measured boundary construct a new boundary to surround an investigating EIT field, and the imaging field can be repartitioned to approximate the actual variant field.
Figure 5 The chest boundary diagram

Full size|PPT slide

3 Experiment

In the finite element method of the EIT imaging field, the investigated imaging area is divided into a number of pixel units. In this paper, the imaging area is partitioned into 812 pixels. In the existing EIT sensors, the thoracic imaging cannot correspond to the actual boundary contour of the human body, since the thoracic profile is arbitrary when breathing. Consequently, the real chest boundary is assumed to be round. With the help of the new sensor system, the real-time variant positions of irregular chest contour have been measured 32 electrodes. The system was evaluated to be stable[14, 15]. Figure 6 shows a data measuring process on a volunteer through the newly designed EIT sensor system.
Figure 6 A real-time measurement to a volunteer

Full size|PPT slide

For the exhalation and inhalation states, measured data is considered to be empty-field data in the EIT imaging process when the inhaling air quantity reaches the maximum. Thus the static measured data is processed as a series of full-field data. The most commonly used in EIT imaging algorithm, linear back projection (LBP)[13], was applied to visually test the imaging effect on the volunteer's chest. In LBP algorithm, the conductivity distribution is assumed to be a number of discrete pixels in the measurement space, so the conductivity in the same pixel region is constant. In fact, the grey value of any pixel can be calculated using a weighted form in LBP algorithm. In this paper, LBP algorithm is used to reconstruct images of actual objectives.
Figure 7 shows a group of reconstructed EIT images of LBP algorithm, where the changing air contents and positions can be observed. In particular, all the reconstructed EIT images are stable and real-time, correctly corresponding to the inhaling and exhaling process. These results validate the effectiveness of the designed sensor system.
Figure 7 A real-time measurement to a volunteer

Full size|PPT slide

4 Summary

The overall system architecture of a new EIT sensor system has been designed and discussed. The purpose of this system is to deal with real-time measurement problems in the case of dynamic change of position in measuring sensors and the investigated fields. Then, some important modules are presented in detail. This method enables the system to adapt to any irregular boundary and shape. The experimental results show the new sensor system is effective. The proposed designing method is simple, general, and efficient, and provides the basis for further application of medical diagnosis and bedside monitoring.

References

1
Demidenko E A, Hartov N S, Paulsen K D. On optimal current patterns for electrical impedance tomography. IEEE Transactions on Bio-Medical Engineering, 2005, 52(2): 238- 248.
2
Dickin F, Wang M. Electrical resistance tomography for processing applications. Measurement Science and Technology, 1999, 7(3): 247- 260.
3
Tien J M, Goldschmidtclermont P J. The case for technobiology: A complement to biotechnology. Journal of Systems Science and Information, 2016, 4(2): 97- 120.
4
Dong F, Xu Y. Application of electrical resistance tomography in two-phase flow measurement. Journal of Engineering Physics and Thermophysics, 2006, 27(5): 791- 794.
5
Marquis F, Coulombe N, Costa R, et al. Electrical impedance tomography's correlation to lung volume is not influenced by anthropometric parameters. Journal of Clinical Monitoring and Computing, 2006, 20(3): 201- 207.
6
Halter R J, Hartov A, Paulsen K D, et al. A broadband high-frequency electrical impedance tomography system for breast imaging. IEEE Transactions on Bio-Medical Engineering, 2008, 55(2): 650- 659.
7
Ma X, Peyton A J, Binns R, et al. Electromagnetic techniques for imaging the cross-section distribution of molten steel flow in the continuous casting nozzle. IEEE Sensors Journal, 2005, 5(2): 224- 232.
8
Dickinson S J, Binns R, Yin W, et al. The development of a multi-frequency electromagnetic instrument for monitoring the phase transformation of hot strip steel. IEEE Transactions on Instrumentation and Measurement, 2005, 56(3): 879- 886.
9
Brown B H, Barber D C, Morice A H, et al. Cardiac and respiratory related electrical impedance changes in the human thorax. IEEE Transactions on Bio-Medical Engineering, 1994, 41(8): 729- 734.
10
Smith R M, Freeston I L, Brown B H, et al. A real-time electrical impedance tomography system for clinical use-design and preliminary results. IEEE Transactions on Bio-Medical Engineering, 1995, 42(2): 133- 140.
11
Adler A, Guardo R, Berthiaume Y, et al. Impedance imaging of lung ventilation: Do we need to account for chest expansion?. IEEE Transactions on Bio-Medical Engineering, 1996, 43(4): 414- 420.
12
Wang B, Ji H, Huang Z, et al. A high-speed data acquisition system for ECT based on the differential sampling method. IEEE Sensors Journal, 2005, 5(2): 308- 312.
13
Yue S, Wu T, Pan J, et al. Fuzzy clustering based ET image fusion. Information Fusion, 2013, 14(4): 487- 497.
14
Wang C, Teng M, Zheng F. Exponential stability and availability analysis of a complex standby system. Journal of Systems Science and Information, 2014, 2(3): 279- 288.
15
Chen L, Wang Q. Prescribed performance adaptive control for a class of non-affine uncertain systems with state and input constraints. Journal of Systems Science and Information, 2016, 4(5): 460- 475.

Acknowledgements

The authors gratefully acknowledge the editor and referees for their insightful comments and helpful suggestions that led to a marked improvement of the article.

Funding

the National Natural Science Foundation of China(61573251)
PDF(578 KB)

132

Accesses

0

Citation

Detail

Sections
Recommended

/