Have a personal or library account? Click to login
From 3D tissue data to impedance using Simpleware ScanFE+IP and COMSOL Multiphysics – a tutorial Cover

From 3D tissue data to impedance using Simpleware ScanFE+IP and COMSOL Multiphysics – a tutorial

Journal of Electrical Bioimpedance
Volume 2 (2011): Issue 1 (January 2011)
By: Fred-Johan Pettersen and  Jan Olav Høgetveit  
Open Access
|Apr 2011

Figures & Tables

Fig. 1

Overview of tutorial. The input data is in DICOM format, and is processed in ScanIP+FE (green). The Model creation and experimenting are done in Multiphysics (blue).
Overview of tutorial. The input data is in DICOM format, and is processed in ScanIP+FE (green). The Model creation and experimenting are done in Multiphysics (blue).

Fig. 2

A typical SIP window. In this particular case, after segmentation.
A typical SIP window. In this particular case, after segmentation.

Fig. 3

Example geometric model.
Example geometric model.

Fig. 4

Example geometric model with interior features shown.
Example geometric model with interior features shown.

Fig. 5

Mesh representation of example geometric model. Coarse mesh.
Mesh representation of example geometric model. Coarse mesh.

Fig. 6

Mesh representation of example geometric model. Fine mesh.
Mesh representation of example geometric model. Fine mesh.

Fig. 7

Electrical properties for the tissues used in this tutorial [8]. The properties are electrical conductivity (σ) plotted in red and relative permittivity (εr) plotted in blue.
Electrical properties for the tissues used in this tutorial [8]. The properties are electrical conductivity (σ) plotted in red and relative permittivity (εr) plotted in blue.

Fig. 8

A typical MPH window with a simulation result in the graphics-pane.
A typical MPH window with a simulation result in the graphics-pane.

Fig. 9

Overview over the segmentation and meshing meshing process. The raw data is coming from a DICOM source such as MRI or CT. As an alternative, one may use pre-segmented data (dashed arrow). This is done using Scan IP+FE, and the output is a mesh usable by Multiphysics.
Overview over the segmentation and meshing meshing process. The raw data is coming from a DICOM source such as MRI or CT. As an alternative, one may use pre-segmented data (dashed arrow). This is done using Scan IP+FE, and the output is a mesh usable by Multiphysics.

Fig. 10

Slice 12 of imported MR data.
Slice 12 of imported MR data.

Fig. 11

Slice 12 of segmented bone.
Slice 12 of segmented bone.

Fig. 12

3D-view of segmented bone.
3D-view of segmented bone.

Fig. 13

Slice 12 of segmented bone and bone marrow. Note that it may be difficult to see that the colours of bone and bone marrow are different in grayscale prints.
Slice 12 of segmented bone and bone marrow. Note that it may be difficult to see that the colours of bone and bone marrow are different in grayscale prints.

Fig. 14

3D-view of bone, bone marrow, and blood.
3D-view of bone, bone marrow, and blood.

Fig. 15

Slice 12 after segmentation.
Slice 12 after segmentation.

Fig. 16

3D-view of thigh after segmentation.
3D-view of thigh after segmentation.

Fig. 17

3D-view of thigh after filtering.
3D-view of thigh after filtering.

Fig. 18

3D-view of thigh with electrodes.
3D-view of thigh with electrodes.

Fig. 19

Overview of modeling process. The inputs are tissue data and the previously generated mesh, which are being merged to create a model. This is the base for our bioimpedance experiment. This is done using Multiphysics.
Overview of modeling process. The inputs are tissue data and the previously generated mesh, which are being merged to create a model. This is the base for our bioimpedance experiment. This is done using Multiphysics.

Fig. 20

3D view of the model with electrodes selected.
3D view of the model with electrodes selected.

Fig. 21

Overview of experiment processes. The process is based on the model previously defined. This is done using Multiphysics.
Overview of experiment processes. The process is based on the model previously defined. This is done using Multiphysics.

Fig. 22

Normal 4-electrode transfer impedance measurement set-up with current stimulus on the outer electrodes and voltage measurement on the inner electrodes. This is a real-world set-up.
Normal 4-electrode transfer impedance measurement set-up with current stimulus on the outer electrodes and voltage measurement on the inner electrodes. This is a real-world set-up.

Fig. 23

Measurement set-up for exploring the principle of reciprocity. This set-up has current stimuli on both outer and inner electrode pairs. This set-up is not used for direct measurements.
Measurement set-up for exploring the principle of reciprocity. This set-up has current stimuli on both outer and inner electrode pairs. This set-up is not used for direct measurements.

Fig. 24

Initial voltage values for simulation. The z-axis is going along the thigh, while the y-axis is going from front to back through the thigh.
Initial voltage values for simulation. The z-axis is going along the thigh, while the y-axis is going from front to back through the thigh.

Fig. 25

Surface plot of potential generated by injecting 1 A through the CC electrodes.
Surface plot of potential generated by injecting 1 A through the CC electrodes.

Fig. 26

Slice plot of potential generated by injecting 1 A through the CC electrodes.
Slice plot of potential generated by injecting 1 A through the CC electrodes.

Fig. 27

Plot of current densities for normal (red) and reciprocal (blue) currents.
Plot of current densities for normal (red) and reciprocal (blue) currents.

Fig. 28

Slice plot of sensitivity inside the thigh.
Slice plot of sensitivity inside the thigh.

Fig. 29

Slice plot of volume impedance density inside the thigh.
Slice plot of volume impedance density inside the thigh.

Tissue mapping table for mapping data between the ScanIP mesh and MPH_

ScanIP nameMPH nameMPH Domain
SkinSkinDry1
FatFatAverageInfiltrated2
MuscleMuscle3
BloodBlood4
BoneBoneCortical5
Bone marrowBoneMarrowInfiltrated6
ElectrodesSteel AISI 43407

Numerical results of our experiment_ The difference between transimp and measured is probably due to the fact that our muscle tissue parameters are not anisotropic as they should have been_

VariableValueDescription
muscle34+0.10itransfer impedance contribution of muscle found by evaluating equation 3 for the complete model.
bone0.014i 0.26−transfer impedance contribution of bone found by evaluating equation 3 for the complete model.
skin0.19i 0.080−transfer impedance contribution of skin found by evaluating equation 3 for the complete model.
all0.013i 40−transfer impedance found by evaluating equation 3 for the complete model.
transimp1.5i 40−transfer impedance found by looking at potential at PU electrodes.
measured30−4.4itransfer impedance measured on the real physical thigh.
DOI: https://doi.org/10.5617/jeb.173 | Journal eISSN: 1891-5469
Journal RSS Feed
Language: English
Page range: 13 - 32
Submitted on: Apr 5, 2011
Published on: Apr 26, 2011
Published by: University of Oslo
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year
Keywords:
bioimpedance,
simulation,
FEM,
COMSOL,
ScanIP,
ScanFE,
transfer impedance,
volume impedance density,
MRI,
CT,
model
Related subjects:
Engineering,
Bioengineering and biomedical engineering,
Biomedical electronics,
Physics,
Spectroscopy and metrology,
Life sciences,
Biophysics,
Medicine,
Biomedical engineering

© 2011 Fred-Johan Pettersen, Jan Olav Høgetveit, published by University of Oslo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Previous articleVolume 2 (2011): Issue 1 (January 2011)