Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Chest EIT:why it’s difficult, and
what we’re doing about it?
EIT 2015Neuchatel, Switzerland 3 June 2015
Andy Adler
Systems and Computer Engineering, Carleton University, Ottawa
Impedance Imaging, Andy Adler, 2015/05/06 1 / 53
Lung Imaging
Source: Kirby et al,Radiology 261.1 (2011)
Pre- and post-salbutamol3He MR images (red)registered to two centercoronal thoracic 1H MRimages (gray scale) for fiverepresentative patients withCOPD
S1, S2: stage II disease,S3, S4: stage III disease,S5: stage IV disease.
Impedance Imaging, Andy Adler, 2015/05/06 2 / 53
Lung Imaging
Source: Kirby et al,Radiology 261.1 (2011)
Pre- and post-salbutamol3He MR images (red)registered to two centercoronal thoracic 1H MRimages (gray scale) for fiverepresentative patients withCOPD
S1, S2: stage II disease,S3, S4: stage III disease,S5: stage IV disease.
Impedance Imaging, Andy Adler, 2015/05/06 2 / 53
Imaging⇒ new clinical insights
Impedance Imaging, Andy Adler, 2015/05/06 3 / 53
Electrical Impedance Tomography
10-day old healthybaby with EITelectrodes
Source: Heinrich, Schiffmann,Frerichs, Klockgether-Radke,Frerichs; Intensive Care Med;2006.eidors3d.sf.net/data contrib/if-neonate-spontaneous
Impedance Imaging, Andy Adler, 2015/05/06 4 / 53
Electrical Impedance Tomography
10-day old healthybaby with EITelectrodes
Source: Heinrich, Schiffmann,Frerichs, Klockgether-Radke,Frerichs; Intensive Care Med;2006.eidors3d.sf.net/data contrib/if-neonate-spontaneous
Impedance Imaging, Andy Adler, 2015/05/06 4 / 53
EIT – Purpose?
Imaging ––Monitoring ++
Impedance Imaging, Andy Adler, 2015/05/06 5 / 53
EIT – Purpose?
Imaging ––
Monitoring ++
Impedance Imaging, Andy Adler, 2015/05/06 5 / 53
EIT – Purpose?
Imaging ––Monitoring ++
Impedance Imaging, Andy Adler, 2015/05/06 5 / 53
EIT – easy?
Audio-frequencyanalog circuitsLow frequency A/DconvertersmV-range voltagemeasurement
Simple imaging algorithms?
Impedance Imaging, Andy Adler, 2015/05/06 6 / 53
EIT – ����easy!
Why is EIT difficult?
Impedance Imaging, Andy Adler, 2015/05/06 7 / 53
Inverse Problems . . . Plato’s cave
Source: iamcriselleeee.files.wordpress.com/2013/11/cave-2.jpg
Impedance Imaging, Andy Adler, 2015/05/06 8 / 53
Plato’s cave . . . Shadows on the wall
Source: iamcriselleeee.files.wordpress.com/2013/11/cave-2.jpg
Impedance Imaging, Andy Adler, 2015/05/06 9 / 53
Inverse Problems
Forward Problem: Forms⇒ Shadows
Inverse Problem: Shadows⇒ Forms
Ill-conditionedSensitivity to some movements is lowIll-posedSome movements don’t change shadowsNoisyFlickering light
Impedance Imaging, Andy Adler, 2015/05/06 10 / 53
Inverse Problems
Forward Problem: Forms⇒ ShadowsInverse Problem: Shadows⇒ Forms
Ill-conditionedSensitivity to some movements is lowIll-posedSome movements don’t change shadowsNoisyFlickering light
Impedance Imaging, Andy Adler, 2015/05/06 10 / 53
Inverse Problems
Forward Problem: Forms⇒ ShadowsInverse Problem: Shadows⇒ Forms
Ill-conditionedSensitivity to some movements is low
Ill-posedSome movements don’t change shadowsNoisyFlickering light
Impedance Imaging, Andy Adler, 2015/05/06 10 / 53
Inverse Problems
Forward Problem: Forms⇒ ShadowsInverse Problem: Shadows⇒ Forms
Ill-conditionedSensitivity to some movements is lowIll-posedSome movements don’t change shadows
NoisyFlickering light
Impedance Imaging, Andy Adler, 2015/05/06 10 / 53
Inverse Problems
Forward Problem: Forms⇒ ShadowsInverse Problem: Shadows⇒ Forms
Ill-conditionedSensitivity to some movements is lowIll-posedSome movements don’t change shadowsNoisyFlickering light
Impedance Imaging, Andy Adler, 2015/05/06 10 / 53
Inverse Problems: EIT
Ill-conditionedLow sensitivity to phenomena ofinterestHigh sensitivity to phenomena not ofinterest
Impedance Imaging, Andy Adler, 2015/05/06 11 / 53
Ill-conditioned . . .
Healthy AdultMale CT slide
Source:eidors3d.sf.net/tutorial/netgen/extrusion
~
~
I~ ~
~ ~
Impedance Imaging, Andy Adler, 2015/05/06 12 / 53
Ill-conditioned . . .
Healthy AdultMale CT slide
Source:eidors3d.sf.net/tutorial/netgen/extrusion
~
~
I~ ~
~ ~
Impedance Imaging, Andy Adler, 2015/05/06 12 / 53
Ill-conditioned . . .
Healthy AdultMale CT slide
Source:eidors3d.sf.net/tutorial/netgen/extrusion
~
~
I
~ ~
~ ~
Impedance Imaging, Andy Adler, 2015/05/06 12 / 53
Ill-conditioned . . .
Healthy AdultMale CT slide
Source:eidors3d.sf.net/tutorial/netgen/extrusion
~
~
I~ ~
~ ~
Impedance Imaging, Andy Adler, 2015/05/06 12 / 53
Ill-conditioned . . .
Healthy AdultMale CT slide
Source:eidors3d.sf.net/tutorial/netgen/extrusion
~
~
I~ ~
~ ~
Impedance Imaging, Andy Adler, 2015/05/06 12 / 53
Finite Element Modelling
Impedance Imaging, Andy Adler, 2015/05/06 13 / 53
Finite Element ModellingStep 1: Create Finite Element Model
%%%% THORAXMDL01% get contoursthorax = shape_library('get','adult_male','boundary');rlung = shape_library('get','adult_male','right_lung');llung = shape_library('get','adult_male','left_lung');% one could also run:% shape_library('get','adult_male');% to get all the info at once in a struct
% show the library imageshape_library('show','adult_male');print_convert thoraxmdl01a.jpg '-density 100'%%%% THORAXMDL02shape = { 1, % height
{thorax, rlung, llung}, % contours[4,50], % perform smoothing with 50 points0.04}; % small maxh (fine mesh)
elec_pos = [ 16, % number of elecs per plane1, % equidistant spacing0.5]'; % a single z-plane
elec_shape = [0.05, % radius0, % circular electrode0.01 ]'; % maxh (electrode refinement)
fmdl = ng_mk_extruded_model(shape, elec_pos, elec_shape);% this similar model is also available as:% fmdl = mk_library_model('adult_male_16el_lungs');
[stim,meas_sel] = mk_stim_patterns(16,1,[0,1],[0,1],{'no_meas_current'}, 1);fmdl.stimulation = stim;
img=mk_image(fmdl,1);img.elem_data(fmdl.mat_idx{2})= 0.3; % rlungimg.elem_data(fmdl.mat_idx{3})= 0.3; % llung
clf; show_fem(img); view(0,70);print_convert thoraxmdl02a.jpg '-density 100'
Impedance Imaging, Andy Adler, 2015/05/06 14 / 53
Finite Element Modelling
Simulated Voltages Voxel Currents
Impedance Imaging, Andy Adler, 2015/05/06 15 / 53
Finite Element ModellingStep 2: Forward solve and current streamlines
img_v = rmfield(img, 'elem_data');img_v.node_data = vh.volt(:,1);img_v.calc_colours.npoints = 128;
PLANE= [inf,inf,0.35]; % show voltages on this slice
subplot(221);show_slices(img_v,PLANE); axis off; axis equalprint_convert thoraxmdl03a.jpg%%%%% THORAXMDL04img_v = img;img_v.fwd_model.mdl_slice_mapper.npx = 64;img_v.fwd_model.mdl_slice_mapper.npy = 64;img_v.fwd_model.mdl_slice_mapper.level = PLANE;q = show_current(img_v, vh.volt(:,1));quiver(q.xp,q.yp, q.xc,q.yc,10,'b');axis tight; axis image; ylim([-1 1]);axis offprint_convert thoraxmdl04a.jpg
%%%% THORAXMDL05img_v.fwd_model.mdl_slice_mapper.npx = 1000;img_v.fwd_model.mdl_slice_mapper.npy = 1000;img_v.fwd_model.mdl_slice_mapper.level = PLANE;
% Calculate at high resolutionq = show_current(img_v, vh.volt(:,1));
pic = shape_library('get','adult_male','pic');imagesc(pic.X, pic.Y, pic.img);% imgt= flipdim(imread('thorax-mdl.jpg'),1); imagesc(imgt);colormap(gray(256)); set(gca,'YDir','normal');hold on
sx = linspace(-.5,.5,15)';sy = 0.05 + linspace(-.5,.5,15)';hh=streamline(q.xp,q.yp, q.xc, q.yc,sx,sy); set(hh,'Linewidth',2, 'color','b');hh=streamline(q.xp,q.yp,-q.xc,-q.yc,sx,sy); set(hh,'Linewidth',2, 'color','b');
axis equal; axis tight; axis off; print_convert thoraxmdl05a.jpg
Impedance Imaging, Andy Adler, 2015/05/06 16 / 53
Thorax Propagation
CT Slice withsimulated currentstreamlines andvoltageequipotentials
~
~
I
~ ~
~ ~
Impedance Imaging, Andy Adler, 2015/05/06 17 / 53
Thorax Propagation
CT Slice withsimulated currentstreamlines andvoltageequipotentials
~
~
I~ ~
~ ~
Impedance Imaging, Andy Adler, 2015/05/06 17 / 53
Changing Conductivity
Heart receivesblood (diastole)and is moreconductive
&%'$
~
~
I~ ~
~ ~
Impedance Imaging, Andy Adler, 2015/05/06 18 / 53
Changing Conductivity
Heart receivesblood (diastole)and is moreconductive
&%'$
~
~
I~ ~
~ ~
Impedance Imaging, Andy Adler, 2015/05/06 19 / 53
Ill-conditioned: low sensitivity
Impedance Imaging, Andy Adler, 2015/05/06 20 / 53
Finite Element Modelling
Step 3: Change conductivity and re-solveselect_fcn = inline('(x-0.1).ˆ2+(y-0.3).ˆ2<0.2ˆ2','x','y','z');memb_frac = elem_select( img.fwd_model, select_fcn);
img=mk_image(fmdl,1);img.elem_data = 1 + memb_frac*2.0;
img.elem_data(fmdl.mat_idx{2})= 0.3; % rlungimg.elem_data(fmdl.mat_idx{3})= 0.3; % llung
show_fem(img)img_v = img;% Stimulate between elecs 16 and 5 to get more interesting patternimg_v.fwd_model.stimulation(1).stim_pattern = sparse([16;5],1,[1,-1],16,1);img_v.fwd_solve.get_all_meas = 1;vh = fwd_solve(img_v);
img_v = rmfield(img, 'elem_data');img_v.node_data = vh.volt(:,1);img_v.calc_colours.npoints = 128;
PLANE= [inf,inf,0.35]; % show voltages on this slice
clfimagesc(pic.X, pic.Y, pic.img); colormap(gray(256)); set(gca,'YDir','normal');hh=streamline(q.xp,q.yp, q.xc, q.yc,sx,sy); set(hh,'Linewidth',2);hh=streamline(q.xp,q.yp,-q.xc,-q.yc,sx,sy); set(hh,'Linewidth',2);
[x y] = meshgrid( linspace(pic.X(1), pic.X(2),size(imgs,1)), ...linspace(pic.Y(2), pic.Y(1),size(imgs,2)));
hold on;contour(x,y,imgs,31);hh= findobj('Type','patch'); set(hh,'LineWidth',2)
hold off; axis off; axis equal; %ylim([50,450]);print_convert thoraxmdl07a.jpg
Impedance Imaging, Andy Adler, 2015/05/06 21 / 53
Non-conductive lungs decrease sensitivity
I+
–
I+
–
Relative sensitivity = 2.8×
Impedance Imaging, Andy Adler, 2015/05/06 22 / 53
Own goal: adjacent stimulation and measurement
I+
–
I+
–
Relative sensitivity = 20.6×
Impedance Imaging, Andy Adler, 2015/05/06 23 / 53
Finite Element Modelling
CODE: Adjacent vs. Skip 4extra={'ball','solid ball = sphere(0,0,1;0.1);'}fmdl= ng_mk_cyl_models(2,[16,1],[.05],extra);fmdl.stimulation = mk_stim_patterns(16,1,[0,1],[0,1],{},1);img= mk_image(fmdl,1); vh = fwd_solve(img);img.elem_data(fmdl.mat_idx{2}) = 2; vi = fwd_solve(img);s1 = norm(vh.meas - vi.meas);
fmdl.stimulation = mk_stim_patterns(16,1,[0,5],[0,5],{},1);img= mk_image(fmdl,1); vh = fwd_solve(img);img.elem_data(fmdl.mat_idx{2}) = 2; vi = fwd_solve(img);s2 = norm(vh.meas - vi.meas);
disp(s2/s1)
Impedance Imaging, Andy Adler, 2015/05/06 24 / 53
Inverse Problems: EIT
Ill-conditionedLow sensitivity to phenomena ofinterestHigh sensitivity to phenomena not ofinterest
Impedance Imaging, Andy Adler, 2015/05/06 25 / 53
EIT Instrumentation
I
300Ω
500Ω
+
–
+5V
–5V
ConductivityChange
(8% diameter)
Impedance Imaging, Andy Adler, 2015/05/06 26 / 53
EIT Instrumentation
I
ConductivityChange
(8% diameter)
300Ω
500Ω
+
–
+5V
–5V
200Ω
0V
2V Common Mode = 1V
Difference = 10μV
Impedance Imaging, Andy Adler, 2015/05/06 27 / 53
Amplifier Requirements
Common Mode Rejection Ratio (CMRR)CMRR = 1V
10µV = 105 = 100 dB
Low Cost Low PowerInstrumentation Amplifier
AD620
Rev. H Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 www.analog.com Fax: 781.326.8703© 2003–2011 Analog Devices, Inc. All rights reserved.
FEATURES Easy to use
Gain set with one external resistor (Gain range 1 to 10,000)
Wide power supply range (±2.3 V to ±18 V) Higher performance than 3 op amp IA designs Available in 8-lead DIP and SOIC packaging Low power, 1.3 mA max supply current
Excellent dc performance (B grade) 50 μV max, input offset voltage 0.6 μV/°C max, input offset drift 1.0 nA max, input bias current 100 dB min common-mode rejection ratio (G = 10)
Low noise 9 nV/√Hz @ 1 kHz, input voltage noise 0.28 μV p-p noise (0.1 Hz to 10 Hz)
Excellent ac specifications 120 kHz bandwidth (G = 100) 15 μs settling time to 0.01%
APPLICATIONS Weigh scales ECG and medical instrumentation Transducer interface Data acquisition systems Industrial process controls Battery-powered and portable equipment
CONNECTION DIAGRAM
–IN
RG
–VS
+IN
RG
+VS
OUTPUT
REF
1
2
3
4
8
7
6
5AD620
TOP VIEW 0077
5-0-
001
Figure 1. 8-Lead PDIP (N), CERDIP (Q), and SOIC (R) Packages
PRODUCT DESCRIPTION
The AD620 is a low cost, high accuracy instrumentation amplifier that requires only one external resistor to set gains of 1 to 10,000. Furthermore, the AD620 features 8-lead SOIC and DIP packaging that is smaller than discrete designs and offers lower power (only 1.3 mA max supply current), making it a good fit for battery-powered, portable (or remote) applications.
The AD620, with its high accuracy of 40 ppm maximum nonlinearity, low offset voltage of 50 μV max, and offset drift of 0.6 μV/°C max, is ideal for use in precision data acquisition systems, such as weigh scales and transducer interfaces. Furthermore, the low noise, low input bias current, and low power of the AD620 make it well suited for medical applications, such as ECG and noninvasive blood pressure monitors.
The low input bias current of 1.0 nA max is made possible with the use of Superϐeta processing in the input stage. The AD620 works well as a preamplifier due to its low input voltage noise of 9 nV/√Hz at 1 kHz, 0.28 μV p-p in the 0.1 Hz to 10 Hz band, and 0.1 pA/√Hz input current noise. Also, the AD620 is well suited for multiplexed applications with its settling time of 15 μs to 0.01%, and its cost is low enough to enable designs with one in-amp per channel.
Table 1. Next Generation Upgrades for AD620 Part Comment AD8221 Better specs at lower price AD8222 Dual channel or differential out AD8226 Low power, wide input range AD8220 JFET input AD8228 Best gain accuracy AD8295 +2 precision op amps or differential out AD8429 Ultra low noise
0 5 10 15 20
30,000
5,000
10,000
15,000
20,000
25,000
0
TOTA
L ER
RO
R, P
PM O
F FU
LL S
CA
LE
SUPPLY CURRENT (mA)
AD620A
RG
3 OP AMP IN-AMP(3 OP-07s)
0077
5-0-
002
Figure 2. Three Op Amp IA Designs vs. AD620
Low Cost Low PowerInstrumentation Amplifier
AD620
Rev. H Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 www.analog.com Fax: 781.326.8703© 2003–2011 Analog Devices, Inc. All rights reserved.
FEATURES Easy to use
Gain set with one external resistor (Gain range 1 to 10,000)
Wide power supply range (±2.3 V to ±18 V) Higher performance than 3 op amp IA designs Available in 8-lead DIP and SOIC packaging Low power, 1.3 mA max supply current
Excellent dc performance (B grade) 50 μV max, input offset voltage 0.6 μV/°C max, input offset drift 1.0 nA max, input bias current 100 dB min common-mode rejection ratio (G = 10)
Low noise 9 nV/√Hz @ 1 kHz, input voltage noise 0.28 μV p-p noise (0.1 Hz to 10 Hz)
Excellent ac specifications 120 kHz bandwidth (G = 100) 15 μs settling time to 0.01%
APPLICATIONS Weigh scales ECG and medical instrumentation Transducer interface Data acquisition systems Industrial process controls Battery-powered and portable equipment
CONNECTION DIAGRAM
–IN
RG
–VS
+IN
RG
+VS
OUTPUT
REF
1
2
3
4
8
7
6
5AD620
TOP VIEW 0077
5-0-
001
Figure 1. 8-Lead PDIP (N), CERDIP (Q), and SOIC (R) Packages
PRODUCT DESCRIPTION
The AD620 is a low cost, high accuracy instrumentation amplifier that requires only one external resistor to set gains of 1 to 10,000. Furthermore, the AD620 features 8-lead SOIC and DIP packaging that is smaller than discrete designs and offers lower power (only 1.3 mA max supply current), making it a good fit for battery-powered, portable (or remote) applications.
The AD620, with its high accuracy of 40 ppm maximum nonlinearity, low offset voltage of 50 μV max, and offset drift of 0.6 μV/°C max, is ideal for use in precision data acquisition systems, such as weigh scales and transducer interfaces. Furthermore, the low noise, low input bias current, and low power of the AD620 make it well suited for medical applications, such as ECG and noninvasive blood pressure monitors.
The low input bias current of 1.0 nA max is made possible with the use of Superϐeta processing in the input stage. The AD620 works well as a preamplifier due to its low input voltage noise of 9 nV/√Hz at 1 kHz, 0.28 μV p-p in the 0.1 Hz to 10 Hz band, and 0.1 pA/√Hz input current noise. Also, the AD620 is well suited for multiplexed applications with its settling time of 15 μs to 0.01%, and its cost is low enough to enable designs with one in-amp per channel.
Table 1. Next Generation Upgrades for AD620 Part Comment AD8221 Better specs at lower price AD8222 Dual channel or differential out AD8226 Low power, wide input range AD8220 JFET input AD8228 Best gain accuracy AD8295 +2 precision op amps or differential out AD8429 Ultra low noise
0 5 10 15 20
30,000
5,000
10,000
15,000
20,000
25,000
0
TOTA
L ER
RO
R, P
PM O
F FU
LL S
CA
LE
SUPPLY CURRENT (mA)
AD620A
RG
3 OP AMP IN-AMP(3 OP-07s)
0077
5-0-
002
Figure 2. Three Op Amp IA Designs vs. AD620
Impedance Imaging, Andy Adler, 2015/05/06 28 / 53
Amplifier Requirements
Common Mode Rejection Ratio (CMRR)CMRR = 1V
10µV = 105 = 100 dB
Low Cost Low PowerInstrumentation Amplifier
AD620
Rev. H Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 www.analog.com Fax: 781.326.8703© 2003–2011 Analog Devices, Inc. All rights reserved.
FEATURES Easy to use
Gain set with one external resistor (Gain range 1 to 10,000)
Wide power supply range (±2.3 V to ±18 V) Higher performance than 3 op amp IA designs Available in 8-lead DIP and SOIC packaging Low power, 1.3 mA max supply current
Excellent dc performance (B grade) 50 μV max, input offset voltage 0.6 μV/°C max, input offset drift 1.0 nA max, input bias current 100 dB min common-mode rejection ratio (G = 10)
Low noise 9 nV/√Hz @ 1 kHz, input voltage noise 0.28 μV p-p noise (0.1 Hz to 10 Hz)
Excellent ac specifications 120 kHz bandwidth (G = 100) 15 μs settling time to 0.01%
APPLICATIONS Weigh scales ECG and medical instrumentation Transducer interface Data acquisition systems Industrial process controls Battery-powered and portable equipment
CONNECTION DIAGRAM
–IN
RG
–VS
+IN
RG
+VS
OUTPUT
REF
1
2
3
4
8
7
6
5AD620
TOP VIEW 0077
5-0-
001
Figure 1. 8-Lead PDIP (N), CERDIP (Q), and SOIC (R) Packages
PRODUCT DESCRIPTION
The AD620 is a low cost, high accuracy instrumentation amplifier that requires only one external resistor to set gains of 1 to 10,000. Furthermore, the AD620 features 8-lead SOIC and DIP packaging that is smaller than discrete designs and offers lower power (only 1.3 mA max supply current), making it a good fit for battery-powered, portable (or remote) applications.
The AD620, with its high accuracy of 40 ppm maximum nonlinearity, low offset voltage of 50 μV max, and offset drift of 0.6 μV/°C max, is ideal for use in precision data acquisition systems, such as weigh scales and transducer interfaces. Furthermore, the low noise, low input bias current, and low power of the AD620 make it well suited for medical applications, such as ECG and noninvasive blood pressure monitors.
The low input bias current of 1.0 nA max is made possible with the use of Superϐeta processing in the input stage. The AD620 works well as a preamplifier due to its low input voltage noise of 9 nV/√Hz at 1 kHz, 0.28 μV p-p in the 0.1 Hz to 10 Hz band, and 0.1 pA/√Hz input current noise. Also, the AD620 is well suited for multiplexed applications with its settling time of 15 μs to 0.01%, and its cost is low enough to enable designs with one in-amp per channel.
Table 1. Next Generation Upgrades for AD620 Part Comment AD8221 Better specs at lower price AD8222 Dual channel or differential out AD8226 Low power, wide input range AD8220 JFET input AD8228 Best gain accuracy AD8295 +2 precision op amps or differential out AD8429 Ultra low noise
0 5 10 15 20
30,000
5,000
10,000
15,000
20,000
25,000
0
TOTA
L ER
RO
R, P
PM O
F FU
LL S
CA
LE
SUPPLY CURRENT (mA)
AD620A
RG
3 OP AMP IN-AMP(3 OP-07s)
0077
5-0-
002
Figure 2. Three Op Amp IA Designs vs. AD620
Low Cost Low PowerInstrumentation Amplifier
AD620
Rev. H Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 www.analog.com Fax: 781.326.8703© 2003–2011 Analog Devices, Inc. All rights reserved.
FEATURES Easy to use
Gain set with one external resistor (Gain range 1 to 10,000)
Wide power supply range (±2.3 V to ±18 V) Higher performance than 3 op amp IA designs Available in 8-lead DIP and SOIC packaging Low power, 1.3 mA max supply current
Excellent dc performance (B grade) 50 μV max, input offset voltage 0.6 μV/°C max, input offset drift 1.0 nA max, input bias current 100 dB min common-mode rejection ratio (G = 10)
Low noise 9 nV/√Hz @ 1 kHz, input voltage noise 0.28 μV p-p noise (0.1 Hz to 10 Hz)
Excellent ac specifications 120 kHz bandwidth (G = 100) 15 μs settling time to 0.01%
APPLICATIONS Weigh scales ECG and medical instrumentation Transducer interface Data acquisition systems Industrial process controls Battery-powered and portable equipment
CONNECTION DIAGRAM
–IN
RG
–VS
+IN
RG
+VS
OUTPUT
REF
1
2
3
4
8
7
6
5AD620
TOP VIEW 0077
5-0-
001
Figure 1. 8-Lead PDIP (N), CERDIP (Q), and SOIC (R) Packages
PRODUCT DESCRIPTION
The AD620 is a low cost, high accuracy instrumentation amplifier that requires only one external resistor to set gains of 1 to 10,000. Furthermore, the AD620 features 8-lead SOIC and DIP packaging that is smaller than discrete designs and offers lower power (only 1.3 mA max supply current), making it a good fit for battery-powered, portable (or remote) applications.
The AD620, with its high accuracy of 40 ppm maximum nonlinearity, low offset voltage of 50 μV max, and offset drift of 0.6 μV/°C max, is ideal for use in precision data acquisition systems, such as weigh scales and transducer interfaces. Furthermore, the low noise, low input bias current, and low power of the AD620 make it well suited for medical applications, such as ECG and noninvasive blood pressure monitors.
The low input bias current of 1.0 nA max is made possible with the use of Superϐeta processing in the input stage. The AD620 works well as a preamplifier due to its low input voltage noise of 9 nV/√Hz at 1 kHz, 0.28 μV p-p in the 0.1 Hz to 10 Hz band, and 0.1 pA/√Hz input current noise. Also, the AD620 is well suited for multiplexed applications with its settling time of 15 μs to 0.01%, and its cost is low enough to enable designs with one in-amp per channel.
Table 1. Next Generation Upgrades for AD620 Part Comment AD8221 Better specs at lower price AD8222 Dual channel or differential out AD8226 Low power, wide input range AD8220 JFET input AD8228 Best gain accuracy AD8295 +2 precision op amps or differential out AD8429 Ultra low noise
0 5 10 15 20
30,000
5,000
10,000
15,000
20,000
25,000
0
TOTA
L ER
RO
R, P
PM O
F FU
LL S
CA
LE
SUPPLY CURRENT (mA)
AD620A
RG
3 OP AMP IN-AMP(3 OP-07s)
0077
5-0-
002
Figure 2. Three Op Amp IA Designs vs. AD620
Impedance Imaging, Andy Adler, 2015/05/06 28 / 53
Amplifier Requirements
Low Cost Low PowerInstrumentation Amplifier
AD620
Rev. H Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 www.analog.com Fax: 781.326.8703© 2003–2011 Analog Devices, Inc. All rights reserved.
FEATURES Easy to use
Gain set with one external resistor (Gain range 1 to 10,000)
Wide power supply range (±2.3 V to ±18 V) Higher performance than 3 op amp IA designs Available in 8-lead DIP and SOIC packaging Low power, 1.3 mA max supply current
Excellent dc performance (B grade) 50 μV max, input offset voltage 0.6 μV/°C max, input offset drift 1.0 nA max, input bias current 100 dB min common-mode rejection ratio (G = 10)
Low noise 9 nV/√Hz @ 1 kHz, input voltage noise 0.28 μV p-p noise (0.1 Hz to 10 Hz)
Excellent ac specifications 120 kHz bandwidth (G = 100) 15 μs settling time to 0.01%
APPLICATIONS Weigh scales ECG and medical instrumentation Transducer interface Data acquisition systems Industrial process controls Battery-powered and portable equipment
CONNECTION DIAGRAM
–IN
RG
–VS
+IN
RG
+VS
OUTPUT
REF
1
2
3
4
8
7
6
5AD620
TOP VIEW 0077
5-0-
001
Figure 1. 8-Lead PDIP (N), CERDIP (Q), and SOIC (R) Packages
PRODUCT DESCRIPTION
The AD620 is a low cost, high accuracy instrumentation amplifier that requires only one external resistor to set gains of 1 to 10,000. Furthermore, the AD620 features 8-lead SOIC and DIP packaging that is smaller than discrete designs and offers lower power (only 1.3 mA max supply current), making it a good fit for battery-powered, portable (or remote) applications.
The AD620, with its high accuracy of 40 ppm maximum nonlinearity, low offset voltage of 50 μV max, and offset drift of 0.6 μV/°C max, is ideal for use in precision data acquisition systems, such as weigh scales and transducer interfaces. Furthermore, the low noise, low input bias current, and low power of the AD620 make it well suited for medical applications, such as ECG and noninvasive blood pressure monitors.
The low input bias current of 1.0 nA max is made possible with the use of Superϐeta processing in the input stage. The AD620 works well as a preamplifier due to its low input voltage noise of 9 nV/√Hz at 1 kHz, 0.28 μV p-p in the 0.1 Hz to 10 Hz band, and 0.1 pA/√Hz input current noise. Also, the AD620 is well suited for multiplexed applications with its settling time of 15 μs to 0.01%, and its cost is low enough to enable designs with one in-amp per channel.
Table 1. Next Generation Upgrades for AD620 Part Comment AD8221 Better specs at lower price AD8222 Dual channel or differential out AD8226 Low power, wide input range AD8220 JFET input AD8228 Best gain accuracy AD8295 +2 precision op amps or differential out AD8429 Ultra low noise
0 5 10 15 20
30,000
5,000
10,000
15,000
20,000
25,000
0
TOTA
L ER
RO
R, P
PM O
F FU
LL S
CA
LE
SUPPLY CURRENT (mA)
AD620A
RG
3 OP AMP IN-AMP(3 OP-07s)
0077
5-0-
002
Figure 2. Three Op Amp IA Designs vs. AD620
Low Cost Low PowerInstrumentation Amplifier
AD620
Rev. H Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 www.analog.com Fax: 781.326.8703© 2003–2011 Analog Devices, Inc. All rights reserved.
FEATURES Easy to use
Gain set with one external resistor (Gain range 1 to 10,000)
Wide power supply range (±2.3 V to ±18 V) Higher performance than 3 op amp IA designs Available in 8-lead DIP and SOIC packaging Low power, 1.3 mA max supply current
Excellent dc performance (B grade) 50 μV max, input offset voltage 0.6 μV/°C max, input offset drift 1.0 nA max, input bias current 100 dB min common-mode rejection ratio (G = 10)
Low noise 9 nV/√Hz @ 1 kHz, input voltage noise 0.28 μV p-p noise (0.1 Hz to 10 Hz)
Excellent ac specifications 120 kHz bandwidth (G = 100) 15 μs settling time to 0.01%
APPLICATIONS Weigh scales ECG and medical instrumentation Transducer interface Data acquisition systems Industrial process controls Battery-powered and portable equipment
CONNECTION DIAGRAM
–IN
RG
–VS
+IN
RG
+VS
OUTPUT
REF
1
2
3
4
8
7
6
5AD620
TOP VIEW 0077
5-0-
001
Figure 1. 8-Lead PDIP (N), CERDIP (Q), and SOIC (R) Packages
PRODUCT DESCRIPTION
The AD620 is a low cost, high accuracy instrumentation amplifier that requires only one external resistor to set gains of 1 to 10,000. Furthermore, the AD620 features 8-lead SOIC and DIP packaging that is smaller than discrete designs and offers lower power (only 1.3 mA max supply current), making it a good fit for battery-powered, portable (or remote) applications.
The AD620, with its high accuracy of 40 ppm maximum nonlinearity, low offset voltage of 50 μV max, and offset drift of 0.6 μV/°C max, is ideal for use in precision data acquisition systems, such as weigh scales and transducer interfaces. Furthermore, the low noise, low input bias current, and low power of the AD620 make it well suited for medical applications, such as ECG and noninvasive blood pressure monitors.
The low input bias current of 1.0 nA max is made possible with the use of Superϐeta processing in the input stage. The AD620 works well as a preamplifier due to its low input voltage noise of 9 nV/√Hz at 1 kHz, 0.28 μV p-p in the 0.1 Hz to 10 Hz band, and 0.1 pA/√Hz input current noise. Also, the AD620 is well suited for multiplexed applications with its settling time of 15 μs to 0.01%, and its cost is low enough to enable designs with one in-amp per channel.
Table 1. Next Generation Upgrades for AD620 Part Comment AD8221 Better specs at lower price AD8222 Dual channel or differential out AD8226 Low power, wide input range AD8220 JFET input AD8228 Best gain accuracy AD8295 +2 precision op amps or differential out AD8429 Ultra low noise
0 5 10 15 20
30,000
5,000
10,000
15,000
20,000
25,000
0
TOTA
L ER
RO
R, P
PM O
F FU
LL S
CA
LE
SUPPLY CURRENT (mA)
AD620A
RG
3 OP AMP IN-AMP(3 OP-07s)
0077
5-0-
002
Figure 2. Three Op Amp IA Designs vs. AD620
AD620
Rev. H | Page 7 of 20
FREQUENCY (Hz)
1000
100
101 10 1000100
CU
RR
ENT
NO
ISE
(fA/
Hz)
0077
5-0-
011
Figure 9. Current Noise Spectral Density vs. Frequency
RTI
NO
ISE
(2.0μV
/DIV
)
TIME (1 SEC/DIV) 0077
5-0-
012
Figure 10. 0.1 Hz to 10 Hz RTI Voltage Noise (G = 1)
RTI
NO
ISE
(0.1μV
/DIV
)
TIME (1 SEC/DIV) 0077
5-0-
013
Figure 11. 0.1 Hz to 10 Hz RTI Voltage Noise (G = 1000)
0077
5-0-
014
Figure 12. 0.1 Hz to 10 Hz Current Noise, 5 pA/Div
100
1000
AD620A
FET INPUTIN-AMP
SOURCE RESISTANCE (Ω)
TOTA
L D
RIF
T FR
OM
25°
C T
O 8
5°C
, RTI
(μV)
100,000
101k 10M
10,000
10k 1M100k
0077
5-0-
015
Figure 13. Total Drift vs. Source Resistance
FREQUENCY (Hz)
CM
R (d
B)
160
01M
80
40
1
60
0.1
140
100
120
100k10k1k10010
G = 1000
G = 100
G = 10
G = 1
20
0077
5-0-
016
Figure 14. Typical CMR vs. Frequency, RTI, Zero to 1 kΩ Source Imbalance
Impedance Imaging, Andy Adler, 2015/05/06 29 / 53
Amplifier Requirements
Low Cost Low PowerInstrumentation Amplifier
AD620
Rev. H Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 www.analog.com Fax: 781.326.8703© 2003–2011 Analog Devices, Inc. All rights reserved.
FEATURES Easy to use
Gain set with one external resistor (Gain range 1 to 10,000)
Wide power supply range (±2.3 V to ±18 V) Higher performance than 3 op amp IA designs Available in 8-lead DIP and SOIC packaging Low power, 1.3 mA max supply current
Excellent dc performance (B grade) 50 μV max, input offset voltage 0.6 μV/°C max, input offset drift 1.0 nA max, input bias current 100 dB min common-mode rejection ratio (G = 10)
Low noise 9 nV/√Hz @ 1 kHz, input voltage noise 0.28 μV p-p noise (0.1 Hz to 10 Hz)
Excellent ac specifications 120 kHz bandwidth (G = 100) 15 μs settling time to 0.01%
APPLICATIONS Weigh scales ECG and medical instrumentation Transducer interface Data acquisition systems Industrial process controls Battery-powered and portable equipment
CONNECTION DIAGRAM
–IN
RG
–VS
+IN
RG
+VS
OUTPUT
REF
1
2
3
4
8
7
6
5AD620
TOP VIEW 0077
5-0-
001
Figure 1. 8-Lead PDIP (N), CERDIP (Q), and SOIC (R) Packages
PRODUCT DESCRIPTION
The AD620 is a low cost, high accuracy instrumentation amplifier that requires only one external resistor to set gains of 1 to 10,000. Furthermore, the AD620 features 8-lead SOIC and DIP packaging that is smaller than discrete designs and offers lower power (only 1.3 mA max supply current), making it a good fit for battery-powered, portable (or remote) applications.
The AD620, with its high accuracy of 40 ppm maximum nonlinearity, low offset voltage of 50 μV max, and offset drift of 0.6 μV/°C max, is ideal for use in precision data acquisition systems, such as weigh scales and transducer interfaces. Furthermore, the low noise, low input bias current, and low power of the AD620 make it well suited for medical applications, such as ECG and noninvasive blood pressure monitors.
The low input bias current of 1.0 nA max is made possible with the use of Superϐeta processing in the input stage. The AD620 works well as a preamplifier due to its low input voltage noise of 9 nV/√Hz at 1 kHz, 0.28 μV p-p in the 0.1 Hz to 10 Hz band, and 0.1 pA/√Hz input current noise. Also, the AD620 is well suited for multiplexed applications with its settling time of 15 μs to 0.01%, and its cost is low enough to enable designs with one in-amp per channel.
Table 1. Next Generation Upgrades for AD620 Part Comment AD8221 Better specs at lower price AD8222 Dual channel or differential out AD8226 Low power, wide input range AD8220 JFET input AD8228 Best gain accuracy AD8295 +2 precision op amps or differential out AD8429 Ultra low noise
0 5 10 15 20
30,000
5,000
10,000
15,000
20,000
25,000
0
TOTA
L ER
RO
R, P
PM O
F FU
LL S
CA
LE
SUPPLY CURRENT (mA)
AD620A
RG
3 OP AMP IN-AMP(3 OP-07s)
0077
5-0-
002
Figure 2. Three Op Amp IA Designs vs. AD620
Low Cost Low PowerInstrumentation Amplifier
AD620
Rev. H Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 www.analog.com Fax: 781.326.8703© 2003–2011 Analog Devices, Inc. All rights reserved.
FEATURES Easy to use
Gain set with one external resistor (Gain range 1 to 10,000)
Wide power supply range (±2.3 V to ±18 V) Higher performance than 3 op amp IA designs Available in 8-lead DIP and SOIC packaging Low power, 1.3 mA max supply current
Excellent dc performance (B grade) 50 μV max, input offset voltage 0.6 μV/°C max, input offset drift 1.0 nA max, input bias current 100 dB min common-mode rejection ratio (G = 10)
Low noise 9 nV/√Hz @ 1 kHz, input voltage noise 0.28 μV p-p noise (0.1 Hz to 10 Hz)
Excellent ac specifications 120 kHz bandwidth (G = 100) 15 μs settling time to 0.01%
APPLICATIONS Weigh scales ECG and medical instrumentation Transducer interface Data acquisition systems Industrial process controls Battery-powered and portable equipment
CONNECTION DIAGRAM
–IN
RG
–VS
+IN
RG
+VS
OUTPUT
REF
1
2
3
4
8
7
6
5AD620
TOP VIEW 0077
5-0-
001
Figure 1. 8-Lead PDIP (N), CERDIP (Q), and SOIC (R) Packages
PRODUCT DESCRIPTION
The AD620 is a low cost, high accuracy instrumentation amplifier that requires only one external resistor to set gains of 1 to 10,000. Furthermore, the AD620 features 8-lead SOIC and DIP packaging that is smaller than discrete designs and offers lower power (only 1.3 mA max supply current), making it a good fit for battery-powered, portable (or remote) applications.
The AD620, with its high accuracy of 40 ppm maximum nonlinearity, low offset voltage of 50 μV max, and offset drift of 0.6 μV/°C max, is ideal for use in precision data acquisition systems, such as weigh scales and transducer interfaces. Furthermore, the low noise, low input bias current, and low power of the AD620 make it well suited for medical applications, such as ECG and noninvasive blood pressure monitors.
The low input bias current of 1.0 nA max is made possible with the use of Superϐeta processing in the input stage. The AD620 works well as a preamplifier due to its low input voltage noise of 9 nV/√Hz at 1 kHz, 0.28 μV p-p in the 0.1 Hz to 10 Hz band, and 0.1 pA/√Hz input current noise. Also, the AD620 is well suited for multiplexed applications with its settling time of 15 μs to 0.01%, and its cost is low enough to enable designs with one in-amp per channel.
Table 1. Next Generation Upgrades for AD620 Part Comment AD8221 Better specs at lower price AD8222 Dual channel or differential out AD8226 Low power, wide input range AD8220 JFET input AD8228 Best gain accuracy AD8295 +2 precision op amps or differential out AD8429 Ultra low noise
0 5 10 15 20
30,000
5,000
10,000
15,000
20,000
25,000
0
TOTA
L ER
RO
R, P
PM O
F FU
LL S
CA
LE
SUPPLY CURRENT (mA)
AD620A
RG
3 OP AMP IN-AMP(3 OP-07s)
0077
5-0-
002
Figure 2. Three Op Amp IA Designs vs. AD620
AD620
Rev. H | Page 7 of 20
FREQUENCY (Hz)
1000
100
101 10 1000100
CU
RR
ENT
NO
ISE
(fA/
Hz)
0077
5-0-
011
Figure 9. Current Noise Spectral Density vs. Frequency
RTI
NO
ISE
(2.0μV
/DIV
)
TIME (1 SEC/DIV) 0077
5-0-
012
Figure 10. 0.1 Hz to 10 Hz RTI Voltage Noise (G = 1)
RTI
NO
ISE
(0.1μV
/DIV
)
TIME (1 SEC/DIV) 0077
5-0-
013
Figure 11. 0.1 Hz to 10 Hz RTI Voltage Noise (G = 1000)
0077
5-0-
014
Figure 12. 0.1 Hz to 10 Hz Current Noise, 5 pA/Div
100
1000
AD620A
FET INPUTIN-AMP
SOURCE RESISTANCE (Ω)TO
TAL
DR
IFT
FRO
M 2
5°C
TO
85°
C, R
TI (μ
V)
100,000
101k 10M
10,000
10k 1M100k
0077
5-0-
015
Figure 13. Total Drift vs. Source Resistance
FREQUENCY (Hz)
CM
R (d
B)
160
01M
80
40
1
60
0.1
140
100
120
100k10k1k10010
G = 1000
G = 100
G = 10
G = 1
20
0077
5-0-
016
Figure 14. Typical CMR vs. Frequency, RTI, Zero to 1 kΩ Source Imbalance
Impedance Imaging, Andy Adler, 2015/05/06 29 / 53
What is contact impedance?
I
zc
zc
z c
Impedance Imaging, Andy Adler, 2015/05/06 30 / 53
What is contact impedance?
I
zc
zc
z c
Impedance Imaging, Andy Adler, 2015/05/06 30 / 53
More challenges
CrosstalkAmplifier input impedance. . .
Question:Geophysicists wont reuse a “hot” electrodefor a few minutes. It becomes polarized.Why do we?
Impedance Imaging, Andy Adler, 2015/05/06 31 / 53
More challenges
CrosstalkAmplifier input impedance. . .
Question:Geophysicists wont reuse a “hot” electrodefor a few minutes. It becomes polarized.Why do we?
Impedance Imaging, Andy Adler, 2015/05/06 31 / 53
Inverse Problems: EIT
Why is EIT difficult?
What are we doing about it?
We = the EIT community
Imaging – –Monitoring ++
Impedance Imaging, Andy Adler, 2015/05/06 32 / 53
Inverse Problems: EIT
Why is EIT difficult?What are we doing about it?
We = the EIT community
Imaging – –Monitoring ++
Impedance Imaging, Andy Adler, 2015/05/06 32 / 53
Inverse Problems: EIT
Why is EIT difficult?What are we doing about it?
We = the EIT community
Imaging – –Monitoring ++
Impedance Imaging, Andy Adler, 2015/05/06 32 / 53
Inverse Problems: EIT
Why is EIT difficult?What are we doing about it?
We = the EIT community
Imaging – –Monitoring ++
Impedance Imaging, Andy Adler, 2015/05/06 32 / 53
What are we doing about it?
Understanding EIT’s sensitivity andlimitationsDeveloping novel measurementprotocols and analysismethodologies
Impedance Imaging, Andy Adler, 2015/05/06 33 / 53
Analysis “pipeline”
EIT Device Raw Measurements
Electrodes
Image reconstruction
Waveforms / ROIs Functionalimage analysis
EIT Measures
R L
CoV: 56%
(Thanks to Inez Frerichs for helping elaborate this concept)
Impedance Imaging, Andy Adler, 2015/05/06 34 / 53
Raw Measurement
EIT Device Raw Measurements
Electrodes
Image reconstruction
Waveforms / ROIs Functionalimage analysis
EIT Measures
R L
CoV: 56%
Where to place electrodes?“Off-plane” contributions?Frame rate?Stimulation / Measurement patterns?
Impedance Imaging, Andy Adler, 2015/05/06 35 / 53
Raw Image Reconstruction
EIT Device Raw Measurements
Electrodes
Image reconstruction
Waveforms / ROIs Functionalimage analysis
EIT Measures
R L
CoV: 56%
Selection of reference/baselineMixed absolute/difference algorithmsElectrode position/movementElectrode errorsData Quality
Impedance Imaging, Andy Adler, 2015/05/06 36 / 53
Reconstruction in Pictures• Forward Problem
Forward Model (linearized)
= ×
Measurements(difference)
Image(difference)
Jacobian
+ noise
• Linear Solution: Minimize norm
Regularized linear inverse model
– ×
Norm weighted bymeasurement accuracy
+ PenaltyFunction
2
Impedance Imaging, Andy Adler, 2015/05/06 37 / 53
Reconstruction in Pictures• Forward Problem
Forward Model (linearized)
= ×
Measurements(difference)
Image(difference)
Jacobian
+ noise
• Linear Solution: Minimize norm
Regularized linear inverse model
– ×
Norm weighted bymeasurement accuracy
+ PenaltyFunction
2
Impedance Imaging, Andy Adler, 2015/05/06 37 / 53
Reconstruction with Data Errors“Traditional”Solution
Electrode errors: “Zero bad data”
– ×
2
σ1
σ2
σ3
σ4
-2
-2
-2
-2
Error HereReplaceWith zero
Error ModelSolution
Electrode Error: Regularized imaging
– ×
2
σ1
σ2
σ3
σ4
-2
-2
-2
-2
Error Here
LowSNRhere
ReplaceWith zero
Impedance Imaging, Andy Adler, 2015/05/06 38 / 53
Reconstruction with Data Errors“Traditional”Solution
Electrode errors: “Zero bad data”
– ×
2
σ1
σ2
σ3
σ4
-2
-2
-2
-2
Error HereReplaceWith zero
Error ModelSolution
Electrode Error: Regularized imaging
– ×
2
σ1
σ2
σ3
σ4
-2
-2
-2
-2
Error Here
LowSNRhere
ReplaceWith zero
Impedance Imaging, Andy Adler, 2015/05/06 38 / 53
Data Quality
Depth Sounder – with analog and digital guages
Impedance Imaging, Andy Adler, 2015/05/06 39 / 53
Data Quality Measure: Concept
Concept: High Quality Data are Consistent
Idea: predict each data point from all others
Originald
⇒ Remove id (i) ⇒ Solve
m(i) ⇒ Predictd (i)
Calculate error
εi = di − d (i)i
Impedance Imaging, Andy Adler, 2015/05/06 40 / 53
Example: Data quality measures(R1–R4 — recruitment: PEEP⇑, T1–T4 — titration: PEEP⇓).
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
R1 R2 R3 R4 T1 T2 T3 T4
No
rmal
ized
dat
a q
ual
ity
valu
es
Protocol steps
Patient1
Patient7
Patient 1
Patient 7
R1 R2 R3 R4 T1 T2 T3 T4
A
B
Clinical data (Wolf, Pediatr Crit Care Med, 2012) and data qualitymetric for each stage of the protocol. A: EIT, B: data quality.
Impedance Imaging, Andy Adler, 2015/05/06 41 / 53
ROIs and Waveforms
EIT Device Raw Measurements
Electrodes
Image reconstruction
Waveforms / ROIs Functionalimage analysis
EIT Measures
R L
CoV: 56%
Lung ROIsAnterioposterior direction
Impedance Imaging, Andy Adler, 2015/05/06 42 / 53
function EIT – fEIT
EIT Device Raw Measurements
Electrodes
Image reconstruction
Waveforms / ROIs Functionalimage analysis
EIT Measures
R L
CoV: 56%
From time sequence – generate image representingspecific physiology
Impedance Imaging, Andy Adler, 2015/05/06 43 / 53
functional EIT (fEIT) images (Tidal Variation)
timeRaw EIT Images
timeEIT
Pix
el
Wav
efor
ms
TV
Standard Deviation
fEIT Image
Tidal Variation
StandardDeviation
Impedance Imaging, Andy Adler, 2015/05/06 44 / 53
functional EIT (fEIT) images (Aeration Change)
timeRaw EIT Images
timeEIT
Pix
el
Wav
efor
ms
ΔEELV
fEIT Image
AerationChange
Impedance Imaging, Andy Adler, 2015/05/06 45 / 53
functional EIT (fEIT) images (Hypo-ventilated lung)
Min Max TV TV
# pi
xels
with
TV
TV
Hypoventilation Map
Hypoventilation
Adequate orgood ventilation
EIT Measure Good or Adequate: 76%Hypoventilated: 24%
timeEIT
Pix
el
Wav
efor
ms
Impedance Imaging, Andy Adler, 2015/05/06 46 / 53
functional EIT (fEIT) images (Hypo-ventilated lung)
Calculate TV and threshold
CODE:path= 'DATA/STUDYNAME/SUBJECT_1/YYYYMMDD/Eit/Viasys/'; dd= dir([path,'*.get']);for idx = 1:length(dd);
vv= eidors_readdata([path,dd(idx).name]);vr = mean(vv,2);img= inv_solve(imdl, vr, vv);img.calc_colours.backgnd=[1,1,1]*.8;ROI0 = ones(1,size(img.elem_data,1));[einsp,eexpi]= find_frc( img, ROI0, 13, '', 2);imgc = calc_slices(img); imgc(imgc==NaN) = 0;TV = imgc(:,:,eexpi) - imgc(:,:,einsp);TV = mean(TV,3).*ROI; % Air is +out = -ROI/2; % blue lungsout( (TV<.2*max(TV(:)))&(ROI==1) ) = 1;out = calc_colours(out, img); out(55:64,:)= [];subplot(3,3,idx); image(out); axis offtext(3,7,dd(idx).name(6:7),'FontSize',16);
end
opt.vert_cut = 10; opt.vert_space = 5;opt.horz_cut = 10; opt.horz_space = 5;print_convert('analyse_step04a.jpg',opt);
Impedance Imaging, Andy Adler, 2015/05/06 47 / 53
fEIT image: Hypo-ventilated lung
c1–c4 — recruitment: PEEP⇑, d1–d4 — titration: PEEP⇓
Impedance Imaging, Andy Adler, 2015/05/06 48 / 53
EIT Measures
EIT Device Raw Measurements
Electrodes
Image reconstruction
Waveforms / ROIs Functionalimage analysis
EIT Measures
R L
CoV: 56%
Goal
Average regional fEIT measuresCharacterizations of the spatial distribution of ventilationExamination-specific measures
Impedance Imaging, Andy Adler, 2015/05/06 49 / 53
#2A Measure of vertical distribution of ventilation
0%
100% CoV is balance point
Measures of the anterioposterior distribution of ventilation1. Center of Ventilation2. Upper to Lower Ventilation Ratio1
1not recommended
Impedance Imaging, Andy Adler, 2015/05/06 50 / 53
EIT
Why is it difficult?What are we doing about it?
Impedance Imaging, Andy Adler, 2015/05/06 51 / 53
EIT Measures
EIT Device Raw Measurements
Electrodes
Image reconstruction
Waveforms / ROIs Functionalimage analysis
EIT Measures
R L
CoV: 56%
Goal
Average regional fEIT measuresCharacterizations of the spatial distribution of ventilationExamination-specific measures
EIT measures using simultaneously measured signalsEIT measures using specific examination
Impedance Imaging, Andy Adler, 2015/05/06 52 / 53
Thank you
Traffic jam on the way to Carleton
Abstract: Currently, the most successful application of EIT is for imaging the thorax, where large movements inconductivity contrasting air and blood can be imaged over time. EIT imaging is difficult due to its low sensitivity tocontrasts deep in the body, because of the diffuse nature of current propagation. EIT is thus sensitive to electrodeproperties, data quality, hardware imperfections and patient movement. To address these issues, several innovativestrategies to analyze and interpret these data have been developed of the past few years. We introduce the conceptof an EIT analysis pipeline, with stages of: measurements, raw images, waveforms, fEIT images, and measures. Wewill discuss recent progress in imaging the chest with EIT, and the image generation and interpretation strategieswhich are required.
Impedance Imaging, Andy Adler, 2015/05/06 53 / 53