PROCEEDINGS OF SPIE
SPIEDigitalLibrary.org/conference-proceedings-of-spie
Development of lung tissue
phantoms for bioluminescent imaging
Durairaj Kumar, Wenxiang Cong, Frank Bohenkamp,
Tarun Kakaday, Peter Taft, et al.
Durairaj Kumar, Wenxiang Cong, Frank Bohenkamp, Tarun Kakaday, Peter
Taft, Lihong V. Wang, Geoffrey McLennan, Eric A. Hoffman, Ge Wang,
"Development of lung tissue phantoms for bioluminescent imaging," Proc.
SPIE 5535, Developments in X-Ray Tomography IV, (26 October 2004); doi:
10.1117/12.560530
Event: Optical Science and Technology, the SPIE 49th Annual Meeting, 2004,
Denver, Colorado, United States
Downloaded From: https://www.spiedigitallibrary.org/conference-proceedings-of-spie on 9/28/2018 Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
Development of lung tissue phanto
ms for bioluminescent imaging
D. Kumar
a
, Cong Wenxiang
a
, Frank Bohenkamp
a
, Tarun Kakaday
c
, Peter Taft
c
, Lihong V. Wang
d
,
McLennan Geoffrey
c
, Eric A. Hoffman, and
∗
Ge Wang
a, b
a
Bioluminescence Tomography Laboratory
b
CT/Micro-CT Laboratory,
Dept. of Radiology,
c
Dept. of Internal Medicine
Univ. of Iowa, Iowa, IA, USA 52242
d
Dept. of Biomedical Engineeri
ng, Texas A&M Univ., TX, USA 77843-3120.
ABSTRACT
White nylon material was chosen to make cylindrical tissue phantoms for development of bioluminescence tomography
techniques. A low-level light source, delivered through a optic fiber of core diameter 200
μ
m, was placed at different
locations on one phantom surface.
The light travels through the phantom, r
eaches the external surface, and is captured
by a liquid nitrogen-cooled CCD camera. The scattering, absorption, and anisotropy parameters of the phantom are
obtained by matching the measured light transmission profiles to the profiles generated by the TracePro software. The
perturbation analysis, with the homogeneous phantoms, demonstr
ated that the imaging system is sufficiently sensitive to
capture intensity change of higher than 0.5nW/cm
2
or a location shift of the light source of more than 200microns. It is
observed that the system can distinguish two point light sources with separation of about 2mm. The perturbation
analysis is also performed with the heterogeneous phantom. Based on our data, we conclude that there is inherent
tomographic information in bioluminescen
t measures taken on the external surf
ace of the mouse, which suggests the
feasibility of bioluminescence tomography for biomedical
research using the small animals, especially the mice.
Keywords
: Bioluminescence tomography (BLT), tissue phantom, optic fiber, light transmission, perturbation analysis.
1. INTRODUCTION
The translucency of the turbid tissue medium has facilitated investigations on biological tissues by optical means. The
non-ionizing nature of the visible region of the spectral radiation further favors various techniques of optical imaging,
such as continuous wave steady state, intensity modulated frequency domain, pulsed time-resolved, optical coherence,
fluorescence and opto-acoustic methods
1
. Following these, bioluminescent imaging has been made possible due to the
recent triumphs of molecular biology an
d the improved sensitivity of the cooled
CCD system. Since the emission is due
to the chemical reaction without any background influen
ce, bioluminescent imaging has attracted more and more
attention
2, 3
. The generated photons propagate from the luminescent
sites to the surface of the
animal. While photons are
quite diffusive and become rather weak at their final destina
tion, they serve as important information providers to depict
neoplastic, dysplastic states of tissue and tumor metastasis
4
. In many cases, bioluminescent imaging is considered more
sensitive than other imaging techniques
5
.
Phantom experiments are required to analyze the photon propagation in various media and structures for developing
bioluminescence tomography. Depending upon the tissue composition, structure and function, various organs such as
the lungs, heart, kidney, brain
etc
. have different tissue optical properties
6, 7
, including the absorption coefficient,
scattering coefficients and anisotropy pa
rameter. Different kinds of phantoms were developed using aqueous solutions
with dyes, wax, resin, gel
etc
., and exploited for evaluating the performance of the imaging systems to observe the
influence of abnormalities in control tissue models
8-12
.
In this paper, we characterize the hom
ogeneous physical phantoms whose optical pr
operties are comparab
le with that of
the lung tissue. To estimate optical parameters of the phan
toms, experimentally measured reflectance or transmission
∗
Corresponding author: ge-wang@uiowa.edu
Developments in X-Ray Tomography IV, edited by Ulrich Bonse, Proc. of SPIE Vol. 5535
(SPIE, Bellingham, WA, 2004) · 0277-786X/04/$15 · doi: 10.1117/12.560530
687
Downloaded From: https://www.spiedigitallibrary.org/conference-proceedings-of-spie on 9/28/2018
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
flux is compared with theoretical predictions based on diffusion theory or Monte Carlo simulation
13
. Also, we report a
perturbation analysis that exam
ines the influence of intensity and displaceme
nt of a single source, as well as the effect
of various separations of paired sources.
2. MATERIALS AND METHODS
2.1 Camera calibration
A system was developed in our laboratory for absolute light intensity calibration and sensitivity measurement of the
nitrogen cooled CCD camera. The system consists of a ca
librated tungsten light source
(LS-1-Cal, Ocean Optics), a
linear variable filter (LVF-H
L, Ocean Optics), a light attenuator (FVA-UV
, Ocean Optics), a bifu
rcated optic fiber
(ZFQ, Ocean Optics) and a spectromete
r (USB-2000 FLG, Ocean Optics). The schematic is shown in Fig. 1.
The camera calibration was done for f-number f/2.8 and magnification 1x. The fiber was kept in front of the camera at a
distance of 15cm away from th
e objective lens of the came
ra. The peak wavelength (
λ
= 650nm) with FWHM 25nm
and arbitrary flux density was selected
(OOIIrrad, Ocean Optics). The camera s
hutter was opened for 0.01second to
obtain the image of the illuminating fiber surface of core
diameter 50micron, and the corresponding maximum pixel
value was noted. The average of 10 such observations was considered as the pixel gray level for the chosen flux density
(
±
2.9
×
10
-17
joules/count), and the same was repeated for various values. The flux density was converted into the number
of photons entering into one pixel (20micron) of the CCD camera in one second. The best fit between flux density and
pixel gray values was obtained by the following linear regression (R
2
= 0.98) equation:
As a result, the pixel values can be associated with the to
tal flux density or number of photons. The sensitivity of the
CCD camera was observed to be
13 photons/pixel-second.
2.2 Parameter estimation
Photons through the optic fiber from the source entered the cylindrical tissue phantom (white nylon) of diameter 20mm
and height 30mm, as shown in Fig. 2.
The CCD camera collected the
transmitted light. To avoid the reentry of light due
to the scattering from the sides of the phantom, it was coated with black.
The Monte Carlo simulation experiments were done using the TracePro software (Lambda Research Corporation,
Littleton, US). It took as input the absorption and scattering
coefficients as well as anisotropy parameter. The various
transmission profiles were obtained with different combinatio
ns of these optical parameters. Using the trial and error
method, the simulated and experimental plots were optimally matched. The optical parameters yielding the best-match
(
χ
2
0.98
) were assigned to the corresponding physical phantom
7, 11, 13
.
2.3 Perturbation analysis
A rectangular slab (50mm
×
70mm
×
20mm) of the tissue phantom was fabricated for a perturbation study. Light guided
through the optic fiber (core diameter 200
μ
m) was delivered on the slab, and we
nt upon the CCD camera for a short
exposure time. The gray level in the image of the illuminating fiber was converted into flux density using the above-
mentioned calibration formula. The following targets were perturbed, respectively.
2
-11
Watts/cm
graylevel
pixel
10
8
density
Flux
×
×
=
Source
S
p
ectromete
r
Filte
r
Attenuato
r
CCD camera
Computer
Bifurcated optic fiber
Fig. 1. Schematic of the system for intens
ity calibration and sensitivity measurement o
f
the nitrogen cooled CCD camera.
688 Proc. of SPIE Vol. 5535
Downloaded From: https://www.spiedigitallibrary.org/conference-proceedings-of-spie on 9/28/2018
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
For a single source with various incident flux densitie
s of small differences, the CCD camera captured the
corresponding transmitted signals. Then, the CCD camera took the outward light from the slab after displacing the
source horizontally step-by-step away from the initial refere
nce position with a given source intensity. Furthermore, the
CCD camera recorded the light signals through two fibers of
initial separation of 0.8mm with
different fixed intensities,
and repeated the measurement for different separa
tion distances between these two optic fibers.
Perturbation studies were also performed using heterogeneous
phantoms with a single source of various flux densities of
small differences and two sources of varying separation.
The heterogeneous phantom was obtained by keeping two
cylindrical phantoms attached to each other having same diameter (30mm) but with different heights viz. 25mm and
5mm whose optical parameters
μ
a
,
μ
s
, and g are 0.1mm
-1
, 3.0mm
-1
and 0.8 as well as 0.20mm
-1
, 0.28mm
-1
and 0.99,
respectively.
The above settings were intended to mimic the presence
bioluminescent source(s) with various intensities and
separations inside the lungs of a small an
imal. Then, student’s t-tests were applied to those data for quantification of the
system sensitivity with respect to source intensity, position and/or separation.
3. RESULTS
Fig. 3 shows the experimental transmission profile of the tissue phantom and the best-matched simulation counterpart,
which is associated with
μ
a
= 0.1mm
-1
,
μ
s
= 3.0mm
-1
and g=0.80. The reported scattering and absorption parameters of
biological tissues vary widely, in the ranges of 1-300cm
-1
and 0.1-10cm
-1
, respectively. Accordingl
y, optical parameters
of the nylon material, from which our lung tissue phan
toms were made, were adjusted to the above regions
1, 6,13
.
Source
(650nm)
CCD
camera
Computer
20mm
30mm
Optic fiber
Fig 2. Schematic of the experimental system to obtain the transmitted signal through
the phantom for estimation of its optical parameters.
D i a m e t e r ( m m )
Transmitted flux
(Watts/m
2
-pixel)
Fig 3. Transmission profiles curve
for a cylindrical tissue phanto
m
(white nylon) and the theoretically best-fit curve by Tracepro with
g
= 0.80,
μ
a
= 0.10cm
-1
and
μ
s
= 3.0cm
-1
.
Proc. of SPIE Vol. 5535 689
Downloaded From: https://www.spiedigitallibrary.org/conference-proceedings-of-spie on 9/28/2018
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
Fig. 4 represents the transmission plots for the physical phantom with various incident source flux densities (1.91, 2.39
and 3.14nW/cm
2
) and the corresponding simulation curves obtained us
ing TracePro. It is observed
that the transmission
signal increases with increment of the
source intensity. A significant difference in transmission corresponds to a
minimum increment in sour
ce intensity of 0.5nW/cm
2
. The significant change (2p<0.05) due to an increment in source
intensity of 0.5nW/cm
2
indicates that about 40 photons would be needed for the detectability after the signal
transmission through the phantom. This is not much worse relative to the sensitivity (about 13 photons) of the CCD
camera itself (
i.e
., exposing it directly to the source).
Fig. 5 presents the physically measured transmission profiles a
ssociated with one source of constant intensity that shifts
horizontally in a small step 0.2mm for
each measurement. The peak of the meas
ured profile shows a clear shift along
with the source movement. Our data reveal that the imaging
system is capable of distingu
ishing a source displacement
larger than 0.2mm, which happens to be the core diameter of the optic fiber.
I n t e n s i t y
( W a t t s / p i x e l )
0
100
200
300
400
500
0.00E+000
2.00E-019
4.00E-019
6.00E-019
8.00E-019
1.00E-018
1.20E-018
1.40E-018
I1
Sim ulation
I2
Sim ulation
I3
Sim ulation
Fig 4. Transmission profiles for the phys
ical phantom with incident source flux
densities of 1.91 (I1), 2.39 (I2
∗
) and 3.14nW/cm
2
(I3
∗
) and their corresponding
simulation profiles by TracePro (
∗
2p<0.05 –Obtained by comparison of I2 and I3
with I1 )
D i s t a n c e ( p i x e l)
D i s t a n c e ( p i x e l)
I n t e n s i t y ( W a t t s / p i x e l )
0
100
200
300
400
500
0.00E+000
1.50E-018
3.00E-018
P1
P2
P3
Fig 5. Transmission profiles from a physical phan
tom for incident source flux density of 3.4nW/cm
2
(
±
0.3nW/cm
2
) at various positions P1 (reference position), P2
∗
(P2=P1+0.2mm), and P3
♠
(P3=P2+0.2mm) (
∗
and
♠
2p<0.05 obtained by comparison of P1
with P2 and P3 respectively).
690 Proc. of SPIE Vol. 5535
Downloaded From: https://www.spiedigitallibrary.org/conference-proceedings-of-spie on 9/28/2018
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
In addition to the single source, the combined effect due to
paired sources is illustrated in Fig. 6. Transmission profiles
were obtained with two sources of separation 0.8mm (D1), 1.00mm (D2), 1.4mm (D3), 1.6mm (D4) and 2.0mm (D5).
The source powers are 8.0±0.5 and 4.0±0.5nW/cm
2
for the bright source on the left and the weak one on the right
respectively, as shown in Fig. 6(a). Comparisons among the measured profiles were made with that corresponding to D1
as the reference, as shown in Fig. 6(b). It is observed that by increasing the horizontal separation between the two
sources, the flux decreases in the left por
tion of the panel, and the larger data points move to the right. The comparisons
D1 vs. D5 and D1 vs. D3 indicate that
the imaging system can detect a source-
separation of about 2mm, albeit a bit
conservatively.
Similar simulation experiments were also performed using the heterogeneous phantoms as given in table 1(a&b). First,
the transmission profiles through the phantom were calculated with source intensity values 10, 11 and 12pW/mm
2
,
respectively. The significance in transmission wa
s obtained with the change in flux of 2pW/mm
2
. The next was to
calculate the transmission profiles with two sources separated by various distances. Their separation ranges were from
0.3mm to 3.4mm with a 0.2mm increment. Two sources were set to have the same flux (1.0pW/mm2). The significance
in transmission was obtained for the separation of 3.4mm. Th
ese may be attributed to the scattering characteristic of the
medium. Further work is in progress to develop more tissue phantoms of various optical parameters for more detailed
analysis.
Fig 6. Perturbation analysis on detectability
of separation of paired sources (a) Various
separation of paired sources. (b) and co
rresponding transmission profiles through the
physical phantom (
∗
and
♠
-2p<0.05 obtained by comparison of D1 with D4 and D5
respectively).
D1
D2
D3
D4
D5
(a)
0
100
200
300
400
500
0. 00E+000
1. 10E-017
2. 20E-017
D1
D2
0. 00E+000
1. 10E-017
2. 20E-017
D1
D3
0. 00E+000
1. 10E-017
2. 20E-017
D1
D4
0. 00E+000
1. 10E-017
2. 20E-017
D1
D5
D i s t a n c e ( p i x e l)
I n t e n s i t y ( W a t t s / p i x e l )
(b)
Proc. of SPIE Vol. 5535 691
Downloaded From: https://www.spiedigitallibrary.org/conference-proceedings-of-spie on 9/28/2018
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
4. CONCLUSION
In conclusion, the physical phantoms have been developed whose optical properties are comparable with that of the
lung tissue. The perturbation analysis has reveal
ed that changes in sour
ce intensity of 0.5nW/mm
2
or source shift of
0.2mm are needed to observe the signifi
cant difference in transmission data. As fa
r as two point sources are concerned,
a separation of 2-3mm is needed to distinguish the two sources directly based on a transmission profile. Our results have
demonstrated that there is definite tomographic informatio
n inherent in bioluminescent data, which ought to be fully
extracted for bioluminescence tomography.
Acknowledgment
This work is partially supported by
the NIH/NIBIB grant 1R21EB001685.
2.43
2.95
11
2.58
2.84
2.65
3.22
12
†
2.21
2.68
10
2.35
2.58
3.10
2.82
Diffused flux
(x10
-9
pW/mm
2
)
Source
Irradiation
(pW/mm
2
)
Table 1a. Diffused flux at the radial
distance of 3mm from the cylindrical
axis for different source irradiation.
(student’s t-test is carried out b
y
comparing the diffused flux at the
radial distance of 3mm for various
source irradiation with the source
irradiation of 10pW/mm
2
)
Table 1b. Diffused flux at the
radial distance of 3mm from the
cylindrical axis for various
sources separartion. (student’s
t-test is carried out b
y
comparing the diffused flux a
t
the radial distance of 3mm fo
r
various source separation with
the sources separation o
f
0.30mm)
7.64
8.28
3.2
9.17
8.03
6.67
7.38
3.4
†
8.79
9.18
0.30
10.70
7.39-21
6.03
7.14
Diffused flux
(x10
-10
pW/mm
2
)
Sources
separation
(mm)
†
-2
p
<0.05.
†
-2
p
<0.05.
692 Proc. of SPIE Vol. 5535
Downloaded From: https://www.spiedigitallibrary.org/conference-proceedings-of-spie on 9/28/2018
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use
REFERENCES
1. Valery Tuchin,
Tissue Optics: Light scattering method
s and instruments for medical diagnosis
, SPIE Press,
Berlingham, Washington, 2000.
2. T. F. Budinger, D. A. Benaron and A. P. Koretsky, “Imaging Transgenic Animals”,
Annu. Rev. Biomed. Eng
.,
1
611-648, 1999.
3. S. R. Cherry, “In vivo molecular and genomic imaging: new challenges for imaging physics”,
Phys. in Med. and
Biol.
,
49
, 13-48, 2004.
4. B. W. Rice, M. D. Cable and M. B. Nelson, “In vivo imaging of light-emitting probes”,
J. of Biomedical Optics
,
6
,
432-440, 2001.
5. Pritha Ray, Anna M. Wu and Sanjiv S. Gambhir, “Optical Bioluminescence and Positron Emission Tomography
Imaging of Novel Fusion reporter
Gene in Tumor Xenografts of Living Mice”,
Cancer Research
,
63
, 1160-1165,
2003.
6. W. F. Cheong, S. A. Prahl and A. J. Welch, “A review of the optical properties of biological tissues”,
IEEE J.
Quant.Electr
.,
26
, 2166-2184, 1990.
7. D. Kumar and M. Singh, “Characterization and imaging of compositional variation in tissues”,
IEEE Trans. on
Biomed. Eng
.,
50
, 1012-1019, 2003.
8. M. Keijzer, W. M. Star and P. R. M. Storchi, “Optical diffusion in layered media.”
Appl. Opt.
,
27
, 1820-24, 1988.
9. J .M. Schmitt, G. X. Zhou, E. C. Walker and R.
T. Wall, “Multilayer model of photon diffusion in skin”.
J. Opt.
Soc. Am.
,
7
, 2141-2153, 1990.
10. Guillermo Marquez, Lihong V Wang, Changjie Wang and
Zhibing Hu, “Development of tissue-simulating optical
phantoms: poly-
N
-isopropylacrylamide solution entrapped inside a hydrogel”,
Phys. in Med. and Biol.
,
44
, 309-318,
1999.
11. R. Srinivasan and M. Singh, “Laser Backscattering an
d Transillumination Imaging of Human Tissues and their
Equivalent Phantoms”,
IEEE Trans. on Biomed. Eng
.,
50
, 724-730, 2003.
12. R. Cubeddu, A. Pifferi, P. Taroni, A. Toricelli and G.
Valentini, “A solid tissue phantom for photon migration
studies”,
Phys. Med. Biol.
42
, 1971-1979, 1997.
13. A.J. Welch and M. J. C. van Gemert,
Optical-Thermal Response of Laser-IrradiatedTissue
, Plenum Press, New
York, 1995.
Proc. of SPIE Vol. 5535 693
Downloaded From: https://www.spiedigitallibrary.org/conference-proceedings-of-spie on 9/28/2018
Terms of Use: https://www.spiedigitallibrary.org/terms-of-use