Video-rate fluorescence diffuse optical tomography for in vivo sentinel lymph node imaging

Abstract

We have developed a fiber-based, video-rate fluorescence diffuse optical tomography (DOT) system for noninvasive in vivo sentinel lymph node (SLN) mapping. Concurrent acquisition of fluorescence and reference signals allowed the efficient generation of ratio-metric data for 3D image reconstruction. Accurate depth localization and high sensitivity to fluorescent targets were established in to depths of >10 mm. In vivo accumulation of indocyanine green (ICG) dye was imaged in the region of the SLN following intradermal injection into the forepaw of rats. These results suggest that video-rate fluorescence DOT has significant potential as a clinical tool for noninvasive mapping of SLN.

Full text

Video-rate fluorescence diffuse optical
tomography for in vivo sentinel lymph node
imaging
Metasebya Solomon,1,2 Brian R. White,2,3 Ralph E. Nothdruft,2 Walter Akers,2 Gail
Sudlow,2 Adam T. Eggebrecht,2 Samuel Achilefu,1,2 and Joseph P. Culver1,2,3
1Department of Biomedical Engineering, Washington University, St. Louis, MO 63110, USA
2Department of Radiology, Washington University School of Medicine, St. Louis, MO 63110, USA
3Department of Physics, Washington University, St. Louis, MO 63110, USA
*culverj@mir.wustl.edu
Abstract: We have developed a fiber-based, video-rate fluorescence diffuse
optical tomography (DOT) system for noninvasive in vivo sentinel lymph
node (SLN) mapping. Concurrent acquisition of fluorescence and reference
signals allowed the efficient generation of ratio-metric data for 3D image
reconstruction. Accurate depth localization and high sensitivity to
fluorescent targets were established in to depths of >10 mm. In vivo
accumulation of indocyanine green (ICG) dye was imaged in the region of
the SLN following intradermal injection into the forepaw of rats. These
results suggest that video-rate fluorescence DOT has significant potential as
a clinical tool for noninvasive mapping of SLN.
©2011 Optical Society of America
OCIS codes: (170.0170) Medical optics and biotechnology; (170.6510) Spectroscopy, tissue
diagnostics; (170.3880) Medical and biological imaging; (170.5270) Photon density waves;
(170.3660) Light propagation in tissues
References and links
1. T. M. Tuttle, M. Colbert, R. Christensen, K. J. Ose, T. Jones, R. Wetherille, J. Friedman, K. Swenson, and K. M.
McMasters, “Subareolar injection of 99mTc facilitates sentinel lymph node identification,” Ann. Surg. Oncol.
9(1), 77–81 (2002).
2. T. Hojo, T. Nagao, M. Kikuyama, S. Akashi, and T. Kinoshita, “Evaluation of sentinel node biopsy by combined
fluorescent and dye method and lymph flow for breast cancer,” Breast 19(3), 210–213 (2010).
3. V. Ntziachristos and R. Weissleder, “Charge-coupled-device based scanner for tomography of fluorescent near-
infrared probes in turbid media,” Med. Phys. 29(5), 803–809 (2002).
4. S. V. Patwardhan, S. R. Bloch, S. Achilefu, and J. P. Culver, “Time-dependent whole-body fluorescence
tomography of probe bio-distributions in mice,” Opt. Express 13(7), 2564–2577 (2005).
5. S. V. Patwardhan and J. P. Culver, “Quantitative diffuse optical tomography for small animals using an ultrafast
gated image intensifier,” J. Biomed. Opt. 13(1), 011009 (2008).
6. V. Ntziachristos and R. Weissleder, “Experimental three-dimensional fluorescence reconstruction of diffuse
media by use of a normalized Born approximation,” Opt. Lett. 26(12), 893–895 (2001).
7. V. Ntziachristos, G. Turner, J. Dunham, S. Windsor, A. Soubret, J. Ripoll, and H. A. Shih, “Planar fluorescence
imaging using normalized data,” J. Biomed. Opt. 10(6), 064007 (2005).
8. J. P. Culver, A. M. Siegel, J. J. Stott, and D. A. Boas, “Volumetric diffuse optical tomography of brain activity,”
Opt. Lett. 28(21), 2061–2063 (2003).
9. M. A. Franceschini and D. A. Boas, “Noninvasive measurement of neuronal activity with near-infrared optical
imaging,” Neuroimage 21(1), 372–386 (2004).
10. B. W. Zeff, B. R. White, H. Dehghani, B. L. Schlaggar, and J. P. Culver, “Retinotopic mapping of adult human
visual cortex with high-density diffuse optical tomography,” Proc. Natl. Acad. Sci. U.S.A. 104(29), 12169–
12174 (2007).
11. G. Marquez, L. V. Wang, S. P. Lin, J. A. Schwartz, and S. L. Thomsen, “Anisotropy in the absorption and
scattering spectra of chicken breast tissue,” Appl. Opt. 37(4), 798–804 (1998).
12. A. M. Nilsson, R. Berg, and S. Andersson-Engels, “Measurements of the optical properties of tissue in
conjunction with photodynamic therapy,” Appl. Opt. 34(21), 4609–4619 (1995).
13. W. F. Cheong, S. A. Prahl, and A. J. Welch, “A Review of the Optical-Properties of Biological Tissues,” IEEE J.
Quantum Electron. 26(12), 2166–2185 (1990).
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3267

14. S. T. Flock, S. L. Jacques, B. C. Wilson, W. M. Star, and M. J. C. van Gemert, “Optical properties of Intralipid: a
phantom medium for light propagation studies,” Lasers Surg. Med. 12(5), 510–519 (1992).
15. A. N. Bashkatov, E. A. Genina, I. V. Korovina, V. I. Kochubey, Yu. P. Sinichkin, and V. V. Tuchin, “In vivo and
in vitro study of control of rat skin optical properties by acting of osmotical liquid,” Proc. SPIE 4224, 300–311
(2000).
16. J. V. Frangioni, S. W. Kim, S. Ohnishi, S. Kim, and M. G. Bawendi, “Sentinel lymph node mapping with type-II
quantum dots,” Methods Mol. Biol. 374, 147–159 (2007).
17. S. T. Proulx, P. Luciani, S. Derzsi, M. Rinderknecht, V. Mumprecht, J. C. Leroux, and M. Detmar, “Quantitative
imaging of lymphatic function with liposomal indocyanine green,” Cancer Res. 70(18), 7053–7062 (2010).
18. R. Sharma, W. Wang, J. C. Rasmussen, A. Joshi, J. P. Houston, K. E. Adams, A. Cameron, S. Ke, S. Kwon, M.
E. Mawad, and E. M. Sevick-Muraca, “Quantitative imaging of lymph function,” Am. J. Physiol. Heart Circ.
Physiol. 292(6), H3109–H3118 (2007).
19. E. M. Sevick-Muraca, R. Sharma, J. C. Rasmussen, M. V. Marshall, J. A. Wendt, H. Q. Pham, E. Bonefas, J. P.
Houston, L. Sampath, K. E. Adams, D. K. Blanchard, R. E. Fisher, S. B. Chiang, R. Elledge, and M. E. Mawad,
“Imaging of lymph flow in breast cancer patients after microdose administration of a near-infrared fluorophore:
feasibility study,” Radiology 246(3), 734–741 (2008).
20. S. L. Troyan, V. Kianzad, S. L. Gibbs-Strauss, S. Gioux, A. Matsui, R. Oketokoun, L. Ngo, A. Khamene, F.
Azar, and J. V. Frangioni, “The FLARE intraoperative near-infrared fluorescence imaging system: a first-in-
human clinical trial in breast cancer sentinel lymph node mapping,” Ann. Surg. Oncol. 16(10), 2943–2952
(2009).
21. L. Sampath, W. Wang, and E. M. Sevick-Muraca, “Near infrared fluorescent optical imaging for nodal staging,”
J. Biomed. Opt. 13(4), 041312 (2008).
22. R. Sharma, J. A. Wendt, J. C. Rasmussen, K. E. Adams, M. V. Marshall, and E. M. Sevick-Muraca, “New
horizons for imaging lymphatic function,” Ann. N. Y. Acad. Sci. 1131(1), 13–36 (2008).
23. B. T. Lee, M. Hutteman, S. Gioux, A. Stockdale, S. J. Lin, L. H. Ngo, and J. V. Frangioni, “The FLARE
intraoperative near-infrared fluorescence imaging system: a first-in-human clinical trial in perforator flap breast
reconstruction,” Plast. Reconstr. Surg. 126(5), 1472–1481 (2010).
24. X. Intes, V. Ntziachristos, J. P. Culver, A. Yodh, and B. Chance, “Projection access order in algebraic
reconstruction technique for diffuse optical tomography,” Phys. Med. Biol. 47(1), N1–N10 (2002).
25. B. W. Pogue, S. P. Poplack, T. O. McBride, W. A. Wells, K. S. Osterman, U. L. Osterberg, and K. D. Paulsen,
“Quantitative hemoglobin tomography with diffuse near-infrared spectroscopy: pilot results in the breast,”
Radiology 218(1), 261–266 (2001).
26. J. P. Culver, R. Choe, M. J. Holboke, L. Zubkov, T. Durduran, A. Slemp, V. Ntziachristos, B. Chance, and A. G.
Yodh, “Three-dimensional diffuse optical tomography in the parallel plane transmission geometry: evaluation of
a hybrid frequency domain/continuous wave clinical system for breast imaging,” Med. Phys. 30(2), 235–247
(2003).
27. R. Choe, A. Corlu, K. Lee, T. Durduran, S. D. Konecky, M. Grosicka-Koptyra, S. R. Arridge, B. J. Czerniecki,
D. L. Fraker, A. DeMichele, B. Chance, M. A. Rosen, and A. G. Yodh, “Diffuse optical tomography of breast
cancer during neoadjuvant chemotherapy: a case study with comparison to MRI,” Med. Phys. 32(4), 1128–1139
(2005).
28. Q. Zhang, T. J. Brukilacchio, A. Li, J. J. Stott, T. Chaves, E. Hillman, T. Wu, M. Chorlton, E. Rafferty, R. H.
Moore, D. B. Kopans, and D. A. Boas, “Coregistered tomographic x-ray and optical breast imaging: initial
results,” J. Biomed. Opt. 10(2), 024033 (2005).
29. D. K. Joseph, T. J. Huppert, M. A. Franceschini, and D. A. Boas, “Diffuse optical tomography system to image
brain activation with improved spatial resolution and validation with functional magnetic resonance imaging,”
Appl. Opt. 45(31), 8142–8151 (2006).
30. T. Austin, A. P. Gibson, G. Branco, R. M. Yusof, S. R. Arridge, J. H. Meek, J. S. Wyatt, D. T. Delpy, and J. C.
Hebden, “Three dimensional optical imaging of blood volume and oxygenation in the neonatal brain,”
Neuroimage 31(4), 1426–1433 (2006).
31. B. W. Zeff, B. R. White, H. Dehghani, B. L. Schlaggar, and J. P. Culver, “Retinotopic mapping of adult human
visual cortex with high-density diffuse optical tomography,” Proc. Natl. Acad. Sci. U.S.A. 104(29), 12169–
12174 (2007).
32. B. R. White, A. Z. Snyder, A. L. Cohen, S. E. Petersen, M. E. Raichle, B. L. Schlaggar, and J. P. Culver,
“Resting-state functional connectivity in the human brain revealed with diffuse optical tomography,”
Neuroimage 47(1), 148–156 (2009).
33. J. P. Culver, V. Ntziachristos, M. J. Holboke, and A. G. Yodh, “Optimization of optode arrangements for diffuse
optical tomography: A singular-value analysis,” Opt. Lett. 26(10), 701–703 (2001).
34. W. Rzyman, O. M. Hagen, R. Dziadziuszko, G. Kobierska-Gulida, A. Karmolinski, M. I. Lothe, W. Paleczka, M.
Murawski, T. Jastrzebski, A. Kopacz, J. Jassem, and J. Skokowski, “Blue-dye intraoperative sentinel lymph node
mapping in early non-small cell lung cancer,” Eur. J. Surg. Oncol. 32(4), 462–465 (2006).
35. L. Cao, M. Breithaupt, and J. Peter, “Geometrical co-calibration of a tomographic optical system with CT for
intrinsically co-registered imaging,” Phys. Med. Biol. 55(6), 1591–1606 (2010).
36. L. Cao and J. Peter, “Bayesian reconstruction strategy of fluorescence-mediated tomography using an integrated
SPECT-CT-OT system,” Phys. Med. Biol. 55(9), 2693–2708 (2010).
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3268

37. A. Li, E. L. Miller, M. E. Kilmer, T. J. Brukilacchio, T. Chaves, J. Stott, Q. Zhang, T. Wu, M. Chorlton, R. H.
Moore, D. B. Kopans, and D. A. Boas, “Tomographic optical breast imaging guided by three-dimensional
mammography,” Appl. Opt. 42(25), 5181–5190 (2003).
38. W. C. Barber, Y. Lin, O. Nalcioglu, J. S. Iwanczyk, N. E. Hartsough, and G. Gulsen, “Combined fluorescence
and x-ray tomography for quantitative in vivo detection of fluorophore,” Technol. Cancer Res. Treat. 9(1), 45–52
(2010).
39. Q. Zhu, T. Durduran, V. Ntziachristos, M. Holboke, and A. G. Yodh, “Imager that combines near-infrared
diffusive light and ultrasound,” Opt. Lett. 24(15), 1050–1052 (1999).
40. Q. Zhu, “Optical tomography with ultrasound localization: initial clinical results and technical challenges,”
Technol. Cancer Res. Treat. 4(3), 235–244 (2005).
41. Q. Zhu, E. B. Cronin, A. A. Currier, H. S. Vine, M. Huang, N. Chen, and C. Xu, “Benign versus malignant breast
masses: optical differentiation with US-guided optical imaging reconstruction,” Radiology 237(1), 57–66 (2005).
1. Introduction
Sentinel lymph node biopsy (SLNB) is the current standard procedure used for prognostic
staging of cancers and therapeutic guidance. SLNB is a minimally-invasive procedure that
involves the removal of sentinel nodes (the first lymph nodes that receive drainage from the
primary tumor) for nodal staging. The location of sentinel lymph nodes (SLNs) is routinely
determined by injecting radioactive lymphophilic tracer dye intra-operatively around the
tumor region. The lymphophilic tracer is commonly composed of a radioactive colloid and/or
optical contrast methylene blue for visual guidance [1,2]. Due to safety issues, a non-invasive
and non-ionizing method for imaging of metastatic lymph nodes and lymphangiogenesis
would be preferred. Fluorescence diffuse optical tomography (DOT) is an emerging deep
tissue (>3 mm) imaging technique that has great potential as an alternative to radioactive
tracer analysis for noninvasive detection and imaging of the sentinel lymph node. While
fluorescence DOT often operates on the time scales of minutes to hours, the method also has
the potential for imaging at higher speeds above the respiratory and cardiac fluctuations,
allowing it to capture pharmacokinetics and pharmacodynamics of diagnostic and therapeutic
agents.
Currently, most fluorescence diffuse optical tomography systems are CCD camera-based
systems that scan at relatively slow speeds (i.e., frame rates < 0.01 Hz). The slow speed is due
in part to the method by which ratio-metric data is acquired. Excitation and emission light
intensity profiles are imaged consecutively, which requires that an interference filter for
blocking the excitation light to be mechanically inserted between the two scans [3–5].
Comparison of sequentially acquired excitation and emission light intensities improves image
quality through the generation of a normalized ratio-metric data set [6,7]. However, the
mechanical requirements of sequential scanning slow data collection and pose a challenge for
real-time imaging. Alternatively, DOT systems developed to image functional absorption
contrast have used time- or frequency-encoding of the illumination with broad spectral
detection to image fast activities (>10 Hz) within the human brain. This style of DOT system
replaces the CCD-camera acquisition with multiple photodiode detectors in order to take
advantage of the high dynamic range and high speed of avalanche photodiodes (APDs) [8–
10]. However, the discrete detector DOT approach has not yet been applied to ratio-metric
data for fluorescence DOT.
Herein, we aim to improve upon current fluorescence DOT platforms by developing a
system with expanded dynamic range and faster data acquisition rates. Our video-rate, fiber-
based fluorescence DOT system design is built upon a previously published APD-based
platform for high-speed DOT [10].
The source-detector grid is designed as a contact probe to pre-operatively identify SLNs
near the tumor region after administration of a fluorescent contrast agent. Biopsy of SLNs
would follow intra-operatively to enable further histological evaluation for the presence of
metastatic cells. These studies with a non-specific, but approved, contrast agent ICG will set
the stage for future work with more specific targeted contrast agents that fluoresce/activate in
the presence of metastatic cells.
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3269

Fiber-based, video-rate fluorescence DOT could additionally improve the flexibility of
imaging by adapting to varying tissue curvatures and performing simultaneous multiple point
illumination and collection, thus increasing the imaging frame rate.
We demonstrate the feasibility of in vivo imaging of the dynamics of dye accumulation in
the region of the sentinel lymph nodes and quantify the accuracy of depth localization and
sensitivity in tissue-simulating phantoms. These results demonstrate that fiber-based, video-
rate fluorescence DOT is a practical and powerful tool that is well suited to a wide array of
potential imaging applications, ranging from sentinel lymph node mapping to monitoring
cancer therapy progression.
2. Methods
The generation of ratiometric (fluorescence-to-reference) data for fluorescence DOT
reconstructions has previously relied on sequentially acquired measurements where the source
laser is kept at the excitation frequency while a band-pass filter in front of the detector
alternates between allowing through excitation and emission (fluorescent) light. The detected
excitation light is then used as a reference to normalize the measured fluorescence. The
acquisition of such ratiometric data allows reconstructions using the normalized Born
approximation, which can be implemented to quantify fluorochrome distribution. However,
while the use of an excitation reference is customary, it is not required. The assumption that
light at the excitation wavelength acts as a good reference inherently assumes similar optical
properties at the two wavelengths. And, if this assumption is made, there is no reason why
light at the fluorescence wavelength could not be used as a reference. With a CCD-based
system, this would be impractical, as there would be no way to determine the difference
between the fluorescent and reference light at the same wavelength. However, with APD-
based detection, high speed allows the use of frequency-encoding regimens to separate out
contributions from different sources. With this encoding scheme, there is no longer any need
for the hardware-based temporal-encoding of switching a filter in and out of the light beam.
This paradigm is the basis for the system presented here.
The fiber-based, video-rate fluorescence DOT system is composed of an alternating grid
of 12 sources and 13 detectors. The source channels contain 785 nm (Thorlabs DL7140-201S)
(3.5 mW) and 830 nm (Thorlabs HL8325G) (0.25 mW) laser diodes, with dedicated laser
diode drivers and control lines for each source, allowing flexible software configurable source
encoding (frequency- and time-encoding) (Fig. 1A). Light from the sources is coupled into 2.5
Fig. 1. Schematic of the video-rate fluorescence DOT hardware. (A) Schematic demonstrating
frequency-encoding of 830 nm and 785 nm laser diode sources. Both reference and excitation
light (at distinct frequencies) are incident on the tissue. Light exiting consists of reference,
excitation, and emission (fluorescent) light. After collimation, the light is passed through a
narrow band optical filter (F) to block the excitation light (785 nm). The resulting detected
light, a sum of the reference transmission (I2) and fluorescence emission (I3), is simultaneously
detected by a single detector. (B) A Fourier transform of the sum of I2 and I3 provides
identification of transmission and emission signals from a single detector.
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3270

mm diameter fiber bundles. The detection channels use optically-filtered discrete avalanche
photodiodes (Hamamatsu C5460-01) digitized with dedicated 24-bit analog-to-digital
converters (MOTU HD 192). A narrowband optical filter with center wavelength 830 +/− 10
nm (CVI) and an out-of-band rejection of OD4 separates the excitation light from fluorescent
and reference signals. An aspheric lens is used to collimate the light in order to optimize the
blocking of excitation light by the narrowband interference filter and enhance fluorescent
signal detection.
This design provides high instantaneous dynamic range (106) and cross-talk rejection
(10−6), so that light levels can be detected over many orders of magnitude. With this scheme,
we acquire frequency-encoded fluorescence emission and reference transmission light levels
simultaneously at each detector through the individual interference filters optimized for
fluorescence emission (Fig. 1A). All data are acquired at a frame rate of 30 Hz. A total of 108
measurements from optode-pairs representing the 1st, 2nd, and 3rd nearest-neighbors are used
for image reconstruction.
2.1. Ratiometric reconstruction
The reference transmission and fluorescent emission light intensities acquired concurrently at
each detector (Fig. 1B) are used to generate ratio-metric data of fluorescence
( )
() ()
,,
emi s i d i emi
rr
φλ
divided by reference transmission
( )
() ()
,,
ref s i d i ref
rr
φλ
. The data are then
reconstructed using the normalized Born approach to correct for tissue and illumination
inhomogeneities, which is written in discrete notation as y = Ax with the following definitions
[6].
In the Normalized Born approach
( )
( )
() ()
() ()
,,
,,
emi s i d i emi
i
ref s i d i ref
rr
yrr
φλ
φλ


=

. When the bleed-through of
the filter is accounted for, then the Normalized Born approach becomes
( ) ( )
( )
() () () ()
() ()
,, , .
,,
emi s i d i emi bleedthrough s i d i
i
ref s i d i ref
rr rr
yrr
φ λφ
φλ

−

=

The ratiometric data (y) is generated by measuring the fluorescence light intensity profile
( )
() ()
,,
emi s i d i emi
rr
φλ
at emission wavelength, emi
λ
, and the reference transmission light intensity
profile,
( )
() ()
,,
ref s i d i ref
rr
φλ
, at the reference wavelength,
ref
λ
. The ith source-detector
measurement (yi) is associated with source (rs(i)) and detector (rd(i)) locations.
The bleed through measurement is made in the absence of a fluorescing agent and
measures the amount of light leakage through the filter. It is determined by
( ) ( ) ( ) ( )
( )
()
()
() () () ()
()
,
, , ,; ;
,
di
exc s i exc
bleedthrough s i d i exc s i exc filter ref s i ref
ref s i ref
r
r r r R r where R r
φλ
φ αφ λ α φ λ φλ
=× = ×× =
α captures the optical density of the filter. In this fiber-based fluorescence DOT system there
are individual detectors and sources, and the sources have separate lasers for both reference
and excitation. Therefore, the α is calculated separately for each source-detector pairs. To
calculate the bleedthrough specific to an imaging study we derived an estimate the
bleedthrough of
emi
φ
based on ref
φ
. Experimentally determined conversion factors (α and R)
are used to account for the difference in power level between the excitation and reference
lasers. R is measured using a power meter at source fiber tips of each sources, and
()di
filter
α
is the
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3271

excitation light measured with and without filter,
( )
()
,
exc s i exc
r
φλ
, for each detectors at the
excitation wavelength ( exc
λ
) of 785 nm.
The sensitivity matrix (Ai,j) is created using the analytic solutions for photon density in the
semi-infinite geometry to obtain the influence of a particular voxel j on every source and
detector measurement i:
( ) ( )
( )
3() ()
,
() ()
,, , , ;
,,
s i j ref j d i emi
o
ij j j
os i d i ref
Gr r Grr
s vh
A xN
DGr r
λλ
λ
=−=∂
The So is calibration factor while h3 represents the voxel volume. The Do and v are the
diffusion coefficient and speed of light in the medium. The two-point Green’s function, G,
models light transport for the given boundary condition. Optical properties at the reference
wavelength of µa = 0.1 cm−1 and µ's = 10 cm−1 were used to model the optical properties of rat
and chicken muscles for both in vivo and phantom studies [11–13]. In our case, a semi-infinite
geometry is implemented.
j
x
is the jth recovered image voxel that contains the fluorescence
yield
j
N∂
obtained using a Moore-Penrose generalized inverse [8].
2.2. Design of phantom studies
To evaluate the performance of the system, different concentrations of ICG targets in 3 mm
diameter plastic tubes were prepared and embedded at 7 mm depth in a breast tissue
mimicking phantom with µa = 0.1 cm−1 and µ's = 10 cm−1 (Fig. 2A). The breast tissue
simulating phantom was constructed by mixing 1% Intralipid solution with Black India ink to
obtain the appropriate absorption and scattering properties [14]. The absorption and reduced
scattering coefficients are chosen based on previously published values of rats and chicken
breast optical properties [11–13,15]. The targets contained solutions of indocyanine green
(Sigma-Aldrich, St Louis, MO) in concentrations ranging from 1 nM to 1 µM. The averaged
values from regions-of-interest (ROIs) of the reconstructed image were compared to known
concentrations to evaluate the sensitivity and linear response of the system.
Analysis of the system’s sensitivity as a function of depth was performed by submerging a
3 mm diameter plastic tube of 4 µM concentration in a tissue mimicking phantom. Starting at
a depth of 5.5 mm, the tube was moved in 1.5 mm increments via a vertical stage until the
middle of the tube reached a depth of 13.5 mm. The depth localization accuracy of the
fluorescent target and sensitivity were calculated by averaging the intensity of the tube ROI
and computing the center-of-mass for the different depths acquired.
2.3. Design of in vivo studies
We conducted a non-invasive preclinical study of fluorescent sentinel lymph node mapping in
rats (n = 5) with our fiber-based, video-rate fluorescence DOT system. The hair was shaved
from the axillary region prior to imaging. In order to test the translational feasibility of the
system to humans (which will require imaging at depths > 1 cm), we also conducted an
experiment mimicking a deeper imaging scenario than would normally be available in a rat by
inserting 8-10 mm of chicken breast tissue between the rat chest wall and the imaging pad
(Fig. 2B). Imaging through 8-10 mm chicken breast increased the depth of the SLNs to 10-12
mm, which is deeper than the typical 2 mm where the SLNs are normally located below the
skin surface in rats. Control images were acquired for 1 minute prior to administration of the
ICG, and the dynamics of the injection in the axillary region were monitored for 10 minutes
continuously.
Animal handling was performed according to the guidelines approved by the Washington
University School of Medicine Animal Studies Committee for humane care and use of
laboratory animals. For in vivo imaging, 50 µL of 100 µM ICG was injected intradermally in
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3272

the left forepaw of 200-250 g female Sprague Dawley rats (HSD, Indianapolis, IN) after
giving a mixture of ketamine (85 mg/kg) and xylazine (15 mg/kg) for anesthesia. After
imaging, rats were then euthanized with an overdose of pentobarbital solution (150 mg/kg, IP)
and the lymph nodes were then resected for verification of ICG uptake.
For reference and verification of ex vivo ICG uptake by the lymph nodes, superficial
fluorescence images were acquired without the layers of chicken breast using the Pearl near-
infrared (NIR) fluorescence imaging system (LiCor Biosciences, Lincoln, NE).
Fig. 2. Schematic of the video-rate fluorescence DOT experimental setup. (A) Schematic of our
experimental setup with an imaging array and a 3 mm ICG tube embedded in a tissue
mimicking phantom. (B) Schematic of the placement of the fiber array on a preclinical animal
model using an 8-10 mm thick chicken breast to simulate a deep tissue imaging situation.
3. Results
3.1. System performance analysis with phantom studies
The Fourier transform of a measurement from a single detector illustrates the use of
frequency-encoding to separate the emission and reference measurements acquired
simultaneously (Fig. 1B). Following frequency-decoding, the intensities of both reference
transmission and emission were retrieved. The data was then used to generate volumetric
reconstructions of the fluorescence distribution. The reconstructed images from various depths
suggest that our video-rate fluorescence DOT system has the capability to accurately section
depths up to 13.5 mm (Fig. 3A). The resolution of the system was tested with simulated data
for 1 × 1 × 1 mm phantom targets at different depths. The size of the reconstructed
perturbation is measured with the full width at half maximum (FWHM) of each target's
reconstruction fluorescence profile. We observe a spatial broadening of the PSF with depth
(Fig. 3B) with resolutions on the order of 12 mm. The center-of-masses (COMs) of the
reconstructed fluorescence tubes (Fig. 3A) were computed by averaging the fluorescence
intensity values for each experimental depth. The linear relationship between the experimental
and the computed COMs demonstrate the depth localization accuracy up to 13.5 mm (Fig.
3C). The sensitivity as a function of depth was computed by taking the mean fluorescence
intensity of the regions-of-interest from reconstructed image for the different depths acquired
(Fig. 3D). The data demonstrate that the signal intensity drops exponentially with depth, as
expected.
The relationship between the true dye concentration and the reconstructed voxel value
(arbitrary units), is shown in Fig. 3E. The graph confirms the high dynamic range and linear
response of our system to varying ICG concentrations as measured by the resulting
fluorescent yield from 1 nM to 1 µM. Values from this phantom analysis are used to generate
a calibration factor for our in vivo studies.
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3273

Fig. 3. System sensitivity analysis with phantom studies. (A) Vertical x-z and y-z slices of
reconstructed experimental data from a fluorescent 3 mm tube target whose center of mass is
located at 7.5 mm, 10.5 mm, and 13.5 mm depths. The system accurately reconstructs the tube
shape with some artifact at the optode positions. (B) Point-spread function analysis using a
simulated image reconstruction. Half-maximum contours of responses for different depths are
shown. (C) Evaluation of the depth localization accuracy of a phantom target. The system has
accurate localization from 6 to 13.5 mm. (D) Sensitivity vs depth. The data demonstrate that
the signal intensity falls off exponentially with depth. (E) The relation between the raw
reconstructed value and the true concentration of the dye was characterized by titration of ICG
from 1 nM to 1 uM concentrations in a 3 mm tube.
3.2. In vivo imaging of the uptake of dye into SLNs
We performed a pre-clinical in vivo study to evaluate the feasibility of imaging sentinel lymph
nodes in rats noninvasively. ICG was injected intradermally into the forpaw and the axillary
region was imaged with DOT. The DOT images of fluorescent lymph dynamics shown at 2
mm were obtained following injection of the ICG (Fig. 4A) (Media 1). Representative
dynamics of fluorescent lymphatic fluid accumulation is obtained from a volume of about 30
mm3 from the region of the sentinel lymph node of the injection site (Fig. 4B). The mean of
the background pixels shows no fluctuation associated with the dynamics of fluorescent dye.
Reflectance fluorescence images were captured using the Pearl NIR fluorescence imaging
system before sacrificing the animal to confirm the DOT results (Fig. 4C). Reflectance
fluorescence images acquired after euthanasia and removal of overlying skin further
confirmed ICG uptake by the lymph node imaged by DOT (inset on Fig. 4D).
We repeated the experiment with an increased imaging depth by inserting 8-10 mm of
chicken breast between the rats (n = 5) and the imaging pad. DOT images of fluorescent
lymph dynamics were acquired and show that high quality imaging can be obtained even at
increased depth (Fig. 5A) (Media 2). Time courses of the SLN region as well as background
pixels demonstrate the representative dynamics of fluorescent lymphatic fluid accumulation
after the injection (in two individual rats Fig. 5B and averaged over all five rats Fig. 5C). The
variability in the lymph dynamics and accumulation in the SLNs can be associated with inter-
and intra-subject variability of the periodic expansion and contraction structures that
surrounds the lymphatic vessels. The results demonstrate the potential of the video-rate
fluorescence DOT system to image the in vivo dynamics of dye accumlation in SLNs.
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3274

Fig. 4. Shallow imaging of lymph dynamics. (A) DOT images of the fluorescence dynamics at
2 mm depth in a rat following injection of ICG into the left forepaw. (B) Time traces of the
dynamics of ICG accumulation in the region of the sentinel lymph node (for comparison the
mean background signal is shown). (C) Reflectance fluorescent imaging of the sentinel lymph
node region demonstrating fluorescence from the injection site (paws) and the lymph vessels
leading to axillary lymph nodes (arrow). (D) Reflectance fluorescent image of the rat after
euthanasia and removal of overlying skin. Inset: fluorescence from ex vivo imaging shows ICG
uptake in the lymph nodes.
4. Discussion
Fiber-based, video-rate fluorescence DOT has the potential to become a powerful and
practical tool for a broad array of imaging applications, ranging from sentinel lymph node
mapping to monitoring cancer therapy progress. We demonstrated the feasibility of a 30 Hz
APD-based fluorescence DOT system. Images of fluorescent targets in tissue mimicking
phantoms were used to confirm the high dynamic range and linear response of the system and
the accurate localization of targets in the range of depths from 5.5 to 13.5 mm. As the optical
properties used in our experiments are higher than those reported for human breast tissues we
expect the depth sensitivity and dynamic range to be improved in human studies.
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3275

Fig. 5. Deep (>10 mm) imaging of lymph dynamics. The video rate fluorescence DOT was
used to image the SLN region in rats through 8-10 mm of chicken breast following ICG
injection into the rat forepaw. (A) The dynamics in a slice at 11 mm depth from a DOT
reconstruction. (B) Time traces of the dynamics (for 2 representative rats) of a region around
the SLN and of the mean of all background pixels. (C) Dynamics of ICG accumulation
averaged over 5 rats.
We also successfully demonstrated the in vivo capabilities of the system by noninvasively
imaging the dynamics of ICG accumulation in a rat. Accumulated dye in the region of the
sentinel lymph nodes was imaged to a depth of 10-12 mm over a 10 minute time course.
These results demonstrate the potential for imaging the pharmacokinetics of fluorescent
diagnostic and therapeutic agents.
Previous two-dimensional fluorescence reflectance imaging (FRI) has been used to image
lymph nodes and the lymphatic systems in animal models and cancer patients [16–22]. For
instance, the lymph dynamics and lymph node images obtained after injection of ICG in
breast cancer patients show the feasibility of implementing FRI for nodal staging [19]. The
specificity of the FRI method was further evaluated with an intraoperative FRI system by
localizing the same SLNs as lymphoscintigraphy in breast cancer patients [20]. This particular
FRI system is currently on clinical trials for intraoperative SLN mapping in cancer patients
[16,17,20,23]. While the FRI method is useful, its sensitivity declines quickly with imaging
depth.
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3276

Three-dimensional DOT methods address several limitations of FRI and can potentially
improve deep-tissue sensitivity, volumetric localization, resolution, and quantitative accuracy.
For example, DOT approaches have been used in small animals to image whole bodies [4,24]
and in humans to image breast [25–28] and brain [29–31] tissues at depths of several
centimeters. One limitation of our current system for application in humans is the limited field
of view. The potentential for an expanded system with up to 48 sources and 48 detectors has
been demonstrated for brain imaging [32]. A higher density imaging array could also
potentially increase the resolution, particularly at the shallower depths [33].
Sentinel lymph node localization is currently performed by administering radiocolloids
conjugated with blue dye for intraoperative guidance [34]. The potential of implementing
other organic optical dyes such as fluorescein and indocyanine green has also been
demonstrated in multiple clinical trials [2,17,19–21,23]. Extensive research is still needed into
targeted optical dyes. Fiber-based, video-rate fluorescence DOT aided with targeted optical
dyes could have the potential to assess the cancer status of SLNs non-invasively in order to
avoid unnecessary surgical procedures.
Multimodal imaging with fluorescence DOT could also play an important role in the
detection and imaging of sentinel lymph nodes for cancer management. DOT could be
combined with, for example, nuclear imaging to improve the accuracy of DOT images during
data processing and image reconstruction [35–38]. In addition to a standalone mode (and due
to the flexibility of the fiber array geometries), video-rate fluorescence DOT also has the
potential to be combined with handheld ultrasound (US) and/or photoacoustic tomography
(PAT). Fibers assembled around the periphery of the PAT/US imaging head could provide
many tracings through the PAT/US imaging volume. A similar approach wherein DOT is
combined with ultrasound has been demonstrated feasible for imaging breast cancer [39–41].
In these scenarios, the fluorescence molecular contrast of DOT would complement the high
resolution function and anatomical data of PAT and US.
5. Conclusion
We have developed a fiber-based, video-rate fluorescence DOT system and demonstrated its
potential for in vivo imaging. We demonstrated that a 30 Hz APD-based DOT system can be
operated in fluorescence mode. A fiber-based imaging array was used to image a fluorescent
target within an agarose phantom and to exhibit our system’s high sensitivity. These results
confirmed the high dynamic range and linear response of the system and accurate localization
of various depths. We have also successfully demonstrated the capability of the system for
sentinel lymph node mapping in rats. Further progress with this technology has the potential
to provide a useful clinical tool for a wide array of imaging applications, ranging from
sentinel lymph node mapping to monitoring cancer therapy progress.
Acknowledgments
This research was supported in part by the Network for Translational Research
U54CA136398 TSP-3 (Culver) and U54CA136398 TSP-1(Achilefu). Dr. Akers is supported
by an award from the National Center for Research Resources (K01RR026095).
#152068 - $15.00 USD
Received 1 Aug 2011; revised 15 Sep 2011; accepted 18 Sep 2011; published 8 Nov 2011
(C) 2011 OSA
1 December 2011 / Vol. 2, No. 12 / BIOMEDICAL OPTICS EXPRESS 3277