Eidors-logo    

EIDORS: Electrical Impedance Tomography and Diffuse Optical Tomography Reconstruction Software

EIDORS (mirror)
Main
Documentation
Tutorials
− Image Reconst
− Data Structures
− Applications
− FEM Modelling
− GREIT
− Old tutorials
− Workshop
Download
Contrib Data
GREIT
Browse Docs
Browse SVN
News
Mailing list
(archive)
FAQ
Developer
                       

 

Hosted by
SourceForge.net Logo

 

GREIT Reconstruction in 3D −

This tutorial shows how to reproduce the 2×16 electrode and 1×32 electrode belt images from the following paper:
Grychtol et al, Focusing EIT reconstructions using two electrode planes p. 17 Conf. EIT 2017, Dartmouth, NH, USA, June 21−24, 2017.
Data for this tutorial are available here: horse-breathing3D2D.mat

Figure: Standing horse with an electrode belt allowing 2×16 electrode and 1×32 electrode belt EIT data recording

GREIT 3D software

Code for Reconstruction using GREIT is in EIDORS v3.9.1. If you have an older version (EIDORS v3.9) you will need the following two files: This code is also very slow to calculate. The actual reconstruction is fast. There are numerous improvements possible, and they're being worked on.

3D FEM for 1×32 electrode belt 

% Model of 32x1 electrode belt
fmdl= ng_mk_ellip_models([4,0.8,1.1,.5],[32,2.0],[0.05]);
% Swisstom BBVet stimulation pattern
skip4 = {32,1,[0,5],[0,5],{'no_meas_current_next1'},1};
[fmdl.stimulation,fmdl.meas_select] = mk_stim_patterns(skip4{:});

3D GREIT reconstruction with 1×32 electrode belt

vopt.imgsz = [32 32];
vopt.square_pixels = true;
vopt.zvec = linspace(-1,1,10)*1.125 + 2;
vopt.save_memory = 1;
opt.noise_figure = 1.0;

% GREIT 3D with a 1x32 electrode layout
[imdl,opt.distr] = GREIT3D_distribution(fmdl, vopt);
imdl2a= mk_GREIT_model(imdl, 0.20, [], opt);

2D GREIT reconstruction with 1×32 electrode belt

% 2D GREIT model
clear opt;
opt.imgsz = [32 32];
opt.square_pixels = true;
opt.noise_figure = 0.5;
img = mk_image(fmdl,1);
imdl2b= mk_GREIT_model(img, 0.25, [], opt);

3D FEM for 2×16 electrode belt 

fmdl= ng_mk_ellip_models([4,0.8,1.1,.5],[16,1.7,2.3],[0.05]);
[fmdl.stimulation,fmdl.meas_select] = mk_stim_patterns(skip4{:});

% "Square" electrode layout
idx = reshape(1:32,2,[])';
idx(2:2:end,:) = fliplr(idx(2:2:end,:));
extraflip= [4:12]; % This belt was made slightly differently
idx(extraflip,:) = fliplr(idx(extraflip,:));
fmdl.electrode(idx) = fmdl.electrode(:);

3D GREIT reconstruction with 2×16 electrode belt

vopt.imgsz = [32 32];
vopt.square_pixels = true;
vopt.zvec = linspace(-1,1,10)*1.125+2;
vopt.save_memory = 1;
opt.noise_figure = 1.0;

% GREIT 3D with 2x16 electrode belt
[imdl,opt.distr] = GREIT3D_distribution(fmdl, vopt);
imdl3= mk_GREIT_model(imdl, 0.20, [], opt);

Load data and reconstruct images

Using data: horse-breathing3D2D.mat 
load horse-breathing3D2D.mat

for i=1:3; switch i;
   case 1; imdl = imdl3;  vv= horse3d; % 2 planes, 3D GREIT
   case 2; imdl = imdl2a; vv= horse2d; % 1 plane, 3D GREIT
   case 3; imdl = imdl2b; vv= horse2d; % 1 plane, 2D GREIT
   end
  
   img = inv_solve(imdl,vv(:,1),vv(:,2:end));
   img.calc_colours.ref_level = 0;
   subplot(121);
   show_slices(img,[inf,inf,2]);

   subplot(122);
   img.elem_data = img.elem_data(:,6);
   if i<3; show_3d_slices(img,[1.6,2.0,2.4],[],[0.4]);
   else;   show_slices(img);
   end;    view(-20,20);

   print_convert(sprintf('GREIT3D_horse06%c.jpg',64+i));
end
Belt3DImage
2×16 ✔
1×32 ✔
1×32

Figure: Reconstructed images of two tidal breaths for two different electrode belts

Last Modified: $Date: 2018-06-14 10:44:51 -0400 (Thu, 14 Jun 2018) $ by $Author: aadler $