Supplementary Information
Electrical Impedance Tomography for Non
-
Invasive Identification of Fatty Liver
Infiltrate in Overweight Individuals
Chih
-
Chiang Chang
1
,4
+
, Zi
-
Yu Huang
2+
, Shu
-
Fu Shih
1,3
, Yuan Luo
2
, Arthur Ko
4
,
Q
ingyu Cui
4
,
Jennifer
Sumner
5
,
Susana Cavallero
4
,
Swarna Das
1
, Wei Gao
2
, Janet Sinsheimer
6
,
7
,
8
, Alex
Bui
1,3
, Jonathan
P.
Jacobs
4
,
9
,
10
, Päivi Pajukanta
7,11
, Holden Wu
1,3
,
Yu
-
Chong
Tai
2
, Zhaoping
Li
4
,
10
,
12
, and Tzung K. Hsiai
1,
2
,
4
,
10
*
1
Department of
Bioengineering, University of California
,
Los Angeles, Los Angeles, CA
2
Department of Medical Engineering, California Institute of Technology, Pasadena, CA
3
Department of Radiological Sciences, David Geffen School of Medicine at UCLA, Los Angeles,
CA
4
Department of Medicine, David Geffen School of Medicine at UCLA, Los Angeles, CA
5
Department of Psychology, College of Life Sciences, UCLA, Los Angeles, CA
6
Department of Biostatistics, Fielding School of Public Hea
l
th, UCLA, Los Angeles, CA
7
Department of
Human Genetics, David Geffen School of Medicine at UCLA, Los Angeles, CA
8
Computational Medicine, David Geffen School of Medicine at UCLA, Los Angeles, CA
9
Division of Digestive Diseases, David Geffen School of Medicine at UCLA, Los Angeles, CA
10
Greater
Los Angeles VA Healthcare System, Los Angeles, CA
11
Institute for Precision Health, David Geffen School of Medicine at UCLA, Los Angeles,
CA
12
Center for Human Nutrition, David Geffen School of Medicine at UCLA, Los Angeles, CA
+Both authors contributed equally.
Supplementary Materials
Fig. S1
.
3
-
D MRI PDFF mapping vs. 3
-
D EIT image
.
Fig. S2
.
Sub
-
analysis of EIT liver conductivity vs. MRI PDFF for all subjects and additional
exclusion of anemic subjects.
Fig. S3
.
Schematic flow of EIT
reconstruction.
Fig. S
4
.
Subject recruitment f
low
chart
.
Table S1
.
Conductivit
i
es
of human tissue
.
Supplementary
Figure
1
Supplementary Figure
1
. 3
-
D MRI PDFF mapping vs. 3
-
D EIT image. (A)
The representative
3
-
D liver boundary condition was established following segmentation of the MRI multi
-
echo
imaging.
(B)
3
-
D MRI PDFF mapping
reveals
a
heterogeneous distribution of MRI PDFF. The red
dashed box highlights the region with a relatively
high
fat fraction.
(C)
3
-
D EIT image unveils the
heterogeneous
gradient of conducti
vi
ty. The dash red box is consistent with that of MRI PDFF
mapping. Thus, the 3
-
D c
omparison between
MRI multi
-
echo imaging and EIT image further
supports the
correlation between MRI fat fraction and EIT
conductivity.
Scale bar: 5 cm.
(A)
(B)
(C)
5 cm
Supplementary Figure
2
Supplementary
Figure
2
.
Sub
-
analysis of
EIT
liver conductivity vs.
MRI PDFF
for all
subjects and additional exclusion of anemic
subjects
.
(A)
The negative correlation between
EIT conductivity and MRI PDFF was reduced to R=
-
0.21 in the presence of preexisting medical
conditions implicated in disturbing tissue electrolytes
(
p
=
0.4
, n
=
1
8
)
.
(B)
The correlation between
EIT liver
vs.
MRI PDFF
was increased to
R =
-
0.7
0 in the absence of anemia subjects (
p
=
0.0049
,
n=
1
4
)
.
The shaded areas reflect the 95% confidence intervals of the linear slopes.
Supplementary
Figure
3
Supplementary Figur
e 3.
Schematic flow of EIT reconstruction. (A)
“Skipping 4” pattern was
used for both current injection and voltage acquisition. There were 4 electrodes separating each
pair of stimulating and detecting electrodes.
(B)
EIT reconstruction was established by solving
the inverse problem via a regularized
Gauss
-
Newton (GN) type solver.
Supplementary
Figure
4
Supplementary Figur
e 4
.
Subject recruitment f
low
chart
.
Supplementary
Ta
ble
1
Conductivit
ies
of human tissue
s at 50 kHz
[
S
m
-
1
].