外文资料-- Simulation Study of Electrode System Optimization in Electrical Impedance Tomography

《外文资料-- Simulation Study of Electrode System Optimization in Electrical Impedance Tomography》由会员分享，可在线阅读，更多相关《外文资料-- Simulation Study of Electrode System Optimization in Electrical Impedance Tomography（4页珍藏版）》请在装配图网上搜索。

1、 Corresponding author:Sha Hong.email:sha_hong_Simulation Study of Electrode System Optimization in Electrical Impedance Tomography Wang Yan,Ren Chaoshi,Zhao Shu,Wang Lei,Sha Hong Institute of Biomedical Engineering Chinese Academy of Medical Sciences&Peking Union Medical College Tianjin,China.E-mail

2、:sarahyky163.cm AbstractElectrode lies in the front of the Electrical impedance tomography(EIT)system and contacts with human body directly.It is one of the most sensitive and the pivotal parts of EIT system.The type and the configuration of electrode system influence the reconstruction image greatl

3、y,especially for the central area of body,where the measurement sensitivity is very low.A simulation platform is developed to optimize the electrode structure parameter based on the coercive equipotential nodes model for the imaging object of the deep area of body.The influences on the measurement s

4、ensitivity and the image reconstruction quality have been studied using different electrode types and configurations.The simulation results suggest that optimal selection of electrode types and configurations can helpfully increase the image quality and measurement sensitivity of the image especiall

5、y in central area.Several type and configuration parameter of EIT electrode that fit for central object imaging are also presented.Keywords-EIT;electrode system;reconstruction quality;measurement sensitivity;simulation evaluation I.0BINTRODUCTION EIT is one of the important research topics in todays

6、 biomedical engineering 1-3.EIT electrode array is one of the most sensitive and the pivotal parts of EIT system.Usually ten or more electrodes are used in EIT system.The electrode array lies in the front of the EIT system and contacts with human body directly.The structure and performance of the el

7、ectrode system greatly influence the original information extraction,real time property of the system,image resolving power,and especially the information extraction from central area,where the EIT measurement sensitivity is poor.In the early optimum study of electrode configuration,it is widely acc

8、epted that current excitation electrodes should be as large as possible while the voltage measuring electrodes are required as small as possible 4.Ping Hua 5 studied the influence of contact impedance between electrode and skin,and showed that the best effect can be obtained when the total width of

9、electrode array is about 80-90%of the imaged areas perimeter.E.J.Woo6 used annular current and voltage electrodes covering 90%of the surface.P.A.T.Pinheiro 7 reported that the optimum area underneath the electrodes is 80%of the overall boundary area for the adjacent patterns and 60%for the opposite

10、ones.Wang Chao et al 8 9 carried on the simulation research on the influence of electrode width using the coercive equipotential node finite element method(FEM)model,and showed that the best effect was obtained when the overlay rate of electrodes is 57.1%.Although continuous progresses on the choice

11、 of electrode structure parameter in EIT have been made,the complete and satisfactory result has not been seen so far.In practice,the design of EIT electrode structure and parameter is still based on experience.Among EIT electrodes,the compound electrode has more complex structure.Four parameters of

12、 the compound electrode restrict each other and their influence on EIT system are much complex.The influences of all parameters should be considered comprehensively.Using only the overlay rate of electrodes is not a good way to deal with the compound electrode optimizing design.In EIT measurement,th

13、e central area,corresponding to deep area of body,is often the region of interest.But the measurement sensitivity of the central area is relatively lower and the image quality is relatively poor.How to raise the measurement sensitivity and improve the image quality of the central area is a key probl

14、em in EIT measurement.The excitation pattern is also an import factor in EIT measurement.Adjacent excitation has more independent measure number.Opposite excitation pattern has less independent measure,but the field distribution is more reasonable than that in adjacent pattern.In this paper,a simula

15、tion study of EIT electrode structure and parameter design aiming at the imaging object of central area is presented.Using the coercive equipotential node model for compound electrode 10,a simulation platform is developed to carry out the optimizing design of the electrode system.Based on the model

16、and the platform,the influences of the electrode number,type,configuration,and excitation pattern of the electrode system on the measurement sensitivity and the image reconstruction quality have been studied.The simulation study of EIT imaging with different depths of imaging object has been made wi

17、th equipotential back projection algorithm.978-1-4244-4713-8/10/$25.00 2010 IEEE II.1BMETHOD A.4BFinite Element Model for Complex Electrodes Electrode plates in round or rectangle shape are usually used in EIT.The projections of both electrodes on the tomogram plane are a line and the length of the

18、line equals to the width of the electrodes,called line electrode,as shown in(1)and(2)of Fig.1.For the compound electrode with a complex structure,the projection on the tomogram plane consists of 3 lines,as shown in(3)and(4)of Fig.1.The excitation electrode consists of two lines,j in Fig.1,on the lef

19、t and the right respectively,and the middle line,c in Fig.1,represents the measurement electrode.There are 4 structure parameters for the compound electrode(Fig.1),the excitation electrode width j,the measurement electrode width c,the space between the excitation electrode and the measurement electr

20、ode b,and the distance between two adjacent compound electrodes a.Each of them is presented by the count of FEM nodes.For an EIT electrode system with certain boundary(such as 256 nodes in this study),when the electrode numbers(32 here)and elements divided in FEM(1210 here)are determined,the sum of

21、electrode wide and the space between adjacent electrodes is fixed too(8 here).Figure 1.The line electrode and compound electrode in EIT A finite element method(FEM)of the compound electrode using coercive equipotential node model has used to study the EIT electrode structure parameter optimizing 10.

22、B.5B Estimate Functions In order to estimate the influences of electrode structure and parameters on the measurement sensitivity and the image reconstruction quality in EIT,the measurement sensitivity function S and the image reconstruction quality function D are defined respectively as follows.1)10

23、BThe measurement sensitivity function S%10011010=NiMjijSNMS Here M is the measurement number corresponding of one current excitation.N is the number of current excitations.Sij denotes the measurement sensitivity for jth measurement in ith current excitation:0101=VVSij Sij shows that the voltage of t

24、he measurement electrode changes from V0 to V1 while the conductance of a certain region in the measured field changes from 0 to 1.S is the average sensitivity of all N times of excitations.It only depends on the electric field distribution due to the conductance change of the measured region,and is

25、 independent of image reconstruction algorithm and hardware errors(suppose that signal to noise ratio of each electrode is fixed).The bigger the S is,the higher the measurement sensitivity is.2)11BThe Image Reconstruction Quality Function D%100255)()(11=MpSpGpGiMD （p=1,2,M）Here M is the number of al

26、l the elements divided in the imaging region.Gs(p)is the gray value of pth element in the imaging model.Gi(p)is the gray value of pth element in the reconstructed image.D shows the normalized average value of the grey scale differences between the reconstructed image and the imaging model of each el

27、ement in the imaging region.D sensitively reflects the reconstructed image quality.Using D,the difference between the reconstructed image and the imaging model can be evaluated without any image processing.The smaller D is,the less the difference between the reconstructed image and the imaging model

28、 is,the higher the reconstructed image quality is.C.6BImaging Models In the study of this paper the total number of finite elements is 1210,including 734 nodes.There are 256 nodes on the circumference of the measured field.6 imaging models corresponding to different measurement depth have been used(

29、Fig.2).The imaged object moves from the boundary to the central area from model 1 to 6,and the model 6 is the imaged object of central area,corresponding to deep area of body.III.2BSIMULATION EXPERIMENT AND RESULT A.7BEffects of Electrode Number EIT system with 8,16,32,64,128 and 256 point electrode

30、s are used respectively for study.The estimate results using functions D for 6 imaging models in Fig.1 is shown in Fig.3.The imaging results of model 3 and model 6 are showed in Fig.4.B.Effects of Electrode Type and Structure Parameter The reconstruction images have been made using the coercive equi

31、potential node method for line electrode and compound electrode.Some comparison studies of the parameter influences on measurement sensitivity and reconstruction quality between line electrode and compound electrode have been carried out.Figure 2.Imaging models of different depths Figure 3.The effec

32、ts of electrode number on image quality Figure 4.Imaging results with different electrode number for model 3 and 6 The total width of both line electrode and compound electrode are set to the same.As an example,the width of line electrode is 5 FEM nodes,while compound electrode has 3 cases,in which

33、the total widths of them are also 5 FEM nodes.The first a=3,j=1,c=3 and b=0,the second a=3,j=1,c=1 and b=1,the third a=3,j=2,c=1 and b=0.The other cases of different width have similar result.The functions S and D of the reconstruction images corresponding to image model 6 and model 3 respectively h

34、ave been calculated and shown in table I and table II.C.9BEffect of Excitation Pattern The effect of excitation pattern of EIT on the reconstructed quality and the measurement sensitivity has investigated.The excitation pattern,which determines how the current flows into the interesting area,influen

35、ces the distribution of electromagnetic field directly.In this study,the adjacent,interval and opposite excitation mode are used.Taken 32 electrodes as an example,the excitation current is applied by two electrodes,and the interval of the electrodes is changed from 1 to 16.When the interval is 1,it

36、is adjacent excitation.And when the interval is 16,it is opposite excitation.Other cases are called interval excitation,Fig.5.The reconstruction results of 32 electrodes at intervals of 1,4,8,12,16 electrodes respectively have been made with the image models shown in Fig.2.The effects of excitation

37、pattern on image quality and measurement sensitivity have been shown in Fig.6 and Fig.7.TABLE I.COMPARISON OF LINE ELECTRODE AND COMPOUND ELECTRODE FOR MODEL 6 TABLE II.COMPARISON OF LINE ELECTRODE AND COMPOUND ELECTRODE FOR MODEL 3 Figure 5.Adjacent excitation,interval excitation and opposite excit

38、ation pattern Figure 6.Effects of excitation pattern on image quality Electrode Type and Structure Parameter D S Line electrode:5 nodes wide 18.706 0.003979 a=3 j=1 c=3 b=0 18.584 0.003897 a=3 j=1 c=1 b=1 18.576 0.003889 a=3 j=2 c=1 b=0 18.589 0.003977 Electrode type and structure parameter D S Line

39、 electrode:5 nodes wide 14.356 0.001871 a=3 j=1 c=3 b=0 14.610 0.001836 a=3 j=1 c=1 b=1 14.475 0.001831 a=3 j=2 c=1 b=0 14.451 0.001875 510152025308163264128256D?electrode countmodel 6model 5model 4model 3model 2model 100.0050.010.0150.020.0251481216S?%?Interval of current electrodesmodel 6model 5mo

40、del 4model 3model 2model 1 Figure 7.Effects of excitation pattern on measurement sensitivity IV.3BRESULTS AND DISCUSSIONS From Fig 3 and Fig 4,it is showed that for noncentral imaging object,smaller electrode number will increase the reconstruction error.But when the electrode number turned up to a

41、certain extent,the improvement became less significant.The image quality is almost the same when electrode number is more than 64 in the study condition.For the imaged object of central area,especially for model 6，reconstruction quality is very similar.When the electrode number enlarged to 128 and 2

42、56,the imaging quality turned a bit worse.This suggests that for central object,bigger electrode number cannot bring better imaging quality.In this paper,the 32 and 64 electrode system is better.In table I,for the imaging object in the central area,compound electrode has higher image quality,than li

43、ne electrode.And in table II,for the imaging object at semi-radius,the effect of line electrompound electrode is helped to improve image quality of the central area.The measurement sensitivity S of line electrode are almost the same with that of the third case of compound electrode,and better than t

44、he other two cases,in both table I and II.It means that,if the total electrode width is fixed,the measurement sensitivity is better when large excitation electrode has been used.From Fig.6 and 7,it can be seen that for every depths of imaging object,image quality turns worse at intervals of bigger d

45、istance,and the measurement sensitivity is increased significantly.It is the suggestion that optimizing of electrode system according to the application requirement,for example the imaging object of different depths and the different excitation pattern,is very important.The simulation study of EIT e

46、lectrode structure and parameter design is helpful to get higher measurement sensitivity and good image quality.12BREFERENCES 1 Barber D.C.and Brown B.H.,“Applied potential tomography”,J.Phsy.E:Sci.instrum.,1984,17,pp.723-733.2 Brown B H,“Medical impedance tomography and process impedance tomography

47、:a brief review”,Meas.Sci.Technol.2001,12,pp.991-996.3 Ren Chaoshi,“Bioimpedance measurement technology.China Medical Device Information”,2004,10,1,pp.21-25.4 Paulson.K.,“Parallel algorithms for three-dimensional electrical impedance tomography”,PhD.Thesis,School of Computing and Mathematical Scienc

48、es,Oxford Brooks University,1992.5 Ping Hua,Eung Je Woo,John G Webster,and Willis J.Tompkins.“Using compound electrodes in electrical impedance tomography”,IEEE Trans.Biome.Engn.,1993,40,1,pp.29-34 6 E.J.Woo,p.Hua,J.G.Webster et al.,“Finite-element method in electrical impedance tomography”.Med.&Bio

49、.Eng.&Comput.1994,32,pp.530-536.7 P.A.T.Pinheiro,W.W.Loh and F.J.Dickin.,“Optimal sized electrodes for electrical resistance tomography”.Electronics Letters 1998,34,1,pp.69-70 8 Wang Chao,Wang Huaxiang,“Optimizing design of electrode construction in medical electrical impedance tomography system”.Jo

50、urnal of the fourth military medical university,2001,22,1,pp.78-80.9 Wang Huaxiang,Wang Chao,Yin Wuliang,“Optimum design of the structure of the electrode for a medical EIT system”,Meas.Sci.Technol.2001,12,pp.1020-1023.10 Wang Yan，Sha Hong,Ren Chaoshi,“Optimum design of electrode structure and parameters in electrical impedance tomography”,Physiol.Meas.2006,27,pp.291306.51015202530351481216D?%?Interval of current electrodesmodel 6model 5model 4model 3model 2model 1