UC Irvine
UC Irvine Previously Published Works
Title
Noninvasive imaging of in vivo blood flow velocity using optical Doppler tomography.
Permalink
https://escholarship.org/uc/item/27s0s83c
Journal
Optics letters, 22(14)
ISSN
0146-9592
Authors
Chen, Z
Milner, TE
Srinivas, S
et al.
Publication Date
1997-07-01
DOI
10.1364/ol.22.001119
Copyright Information
This work is made available under the terms of a Creative Commons Attribution License,
availalbe at https://creativecommons.org/licenses/by/4.0/
Peer reviewed
eScholarship.org Powered by the California Digital Library
University of California
July 15, 1997 / Vol. 22, No. 14 / OPTICS LETTERS 1119
Noninvasive imaging of in vivo blood flow velocity using
optical Doppler tomography
Zhongping Chen, Thomas E. Milner, Shyam Srinivas, Xiaojun Wang, Arash Malekafzali,
Martin J. C. van Gemert, and J. Stuart Nelson
Beckman Laser Institute and Medical Clinic, University of California, Irvine, Irvine, California 92612
Received March 3, 1997
We report the development of an optical technique for noninvasive imaging of in vivo blood f low dynamics and
tissue structures with high spatial resolution (2–10 mm) in biological systems. The technique is based on
optical Doppler tomography (ODT), which combines Doppler velocimetry with optical coherence tomography
to measure blood flow velocity at discrete spatial locations. The exceptionally high resolution of ODT
permits noninvasive in vivo imaging of both blood microcirculation and tissue structures surrounding the
vessel, which has significance for biomedical research and clinical applications. Tomographic imaging of
in vivo blood flow velocity in the chick chorioallantoic membrane and in rodent skin is demonstrated. 1997
Optical Society of America
High-resolution noninvasive techniques for in vivo
blood f low imaging are not currently available as a
diagnostic tool in medicine. Such a technique could
have a significant impact on biomedical research and
clinical diagnosis.
1
Numerous approaches have been
investigated, including Doppler ultrasound,
2
conven-
tional angiography,
3
laser Doppler flowmetry (LDF),
4
and magnetic-resonance angiography.
3
Each of these
techniques has limitations. Conventional LDF, for ex-
ample, has been used to measure mean blood perfusion
in the peripheral microcirculation. However, strong
optical scattering in biological tissue limits spatially
resolved f low measurements by LDF. Although
Doppler ultrasound imaging provides a means to
resolve f low velocities at different locations in a scat-
tering medium, the relatively long acoustic wavelength
required for deep tissue penetration limits spatial
resolution to ,200 mm. Coherence techniques based
on detection of the interference-fringe intensity of
light backscattered from a sample have been used
to image biological samples with micrometer resolu-
tion.
5–7
Consequently, f low-velocity sensors based on
the coherence interferometry technique are expected
to improve spatial resolution.
8,9
In this Letter we
report what is believed to be the first noninvasive
in vivo imaging of blood flow by use of optical Doppler
tomography (ODT). Compared with other microvas-
cular imaging techniques, ODT is noninvasive and
noncontact, has high spatial resolution (2–10 mm),
and provides simultaneous information regarding not
only in vivo blood f low at discrete locations but also
the tissue structure surrounding the vessel.
ODT combines LDF
4
with optical coherence tomog-
raphy,
5– 7,10,11
which uses a Michelson interferometer
with a low-coherence light source to obtain sectional
images of biological materials. In ODT (Fig. 1), the
sample and the reference mirrors constitute the two
arms of an interferometer. One determines the am-
plitude of backscattered light from the sample by
measuring the interference-fringe intensity generated
between reference and target beams. High axial spa-
tial resolution is possible because light backscattered
from the sample recombines with the reference beam in
the 2 3 2 coupler and interferes only when the optical-
path-length difference is within the coherence length
of the source light. When light backscattered from a
moving particle interferes with a reference beam, beat-
ing at the Doppler frequency (f
D
) occurs, shifting the
frequency of the interference-fringe intensity from that
of the optical phase modulation by f
D
:
f
D
1
2p
sk
s
2 k
i
d ? V , (1)
where k
i
and k
s
are wave vectors of incoming and
scattered light, respectively, and V is the velocity
vector of the moving particles.
Fig. 1. Schematic diagram of the ODT system. Light
emitted from the superluminescent diode sl
0
850 nm,
Dl
FWHM
25 nmd is coupled into a single-mode fiber and
split into reference and target arms by a 2 3 2 (50/50)
fiber coupler. Stress birefringence is used to match the
polarization of reference and sample beams and to optimize
fringe contrast. The optical path lengths of light in the
reference and the target arms are modulated (1600 Hz)
with piezoelectric cylinders electrically driven by a ramp
waveform. The target arm is tilted 70–75
±
with respect to
the direction of flow. Optical interference-fringe intensity
is measured by a photodetector and digitized with a
16-bit analog–digital converter. Structural and f low-
velocity images are calculated from the digitized fringe
intensities.
0146-9592/97/141119-03$10.00/0 1997 Optical Society of America
1120 OPTICS LETTERS / Vol. 22, No. 14 / July 15, 1997
Fluid-flow velocity at each pixel is determined by
measurement of the Doppler frequency shift, which
is defined as the difference between the carrier fre-
quency established by the optical phase modulation
and the centroid of the measured power spectrum at
each pixel. Two-dimensional images are formed by
sequential lateral scans at a constant horizontal veloc-
ity of 800 mmys, followed by incremental probe move-
ments (10 mm) in the vertical (axial) direction. In
addition to the flow-velocity image, one can obtain a
tomographic structural image from the same scan si-
multaneously by calculating at each pixel the magni-
tude of the power spectrum at the carrier frequency
established by the optical phase modulation.
To demonstrate the ability of ODT to image in vivo
blood f low, we studied two biological models, the chick
chorioallantoic membrane (CAM) and rodent skin.
The CAM is a well-established model for studying
microvasculature and has been used extensively to in-
vestigate the effects of vasoactive drugs as well as opti-
cal and thermal processes in blood vessels.
12
Because
the CAM microvasculature is located in a transparent
matrix, direct viewing and noninvasive imaging of the
blood vessels are possible after the apex of the chick
egg shell is removed. The effects of optical scattering
in biological tissues are investigated in the rodent skin
model.
In the CAM model, blood flow in a vein is imaged to
minimize the effect of pulsation. Structural (Fig. 2A)
and velocity (Fig. 2B) images of CAM blood f low are ob-
tained simultaneously. The blood vessel wall, chorion
membrane, and yolk sac membrane are evident in the
structural image. In the velocity image, static regions
(V 0) in the CAM appear dark, and blood moving at
different velocities is evident. The magnitude of blood
flow velocity is maximal at the vessel center and de-
creases monotonically toward the peripheral wall. A
horizontal cross section of the velocity image near the
vessel center is shown in Fig. 2C. The excellent fit
of the velocity profile to a parabolic function indicates
that blood flow in the CAM vein is laminar.
In the rodent (Sprague-Dawley) skin model, in vivo
blood f low in both veins and arteries is imaged by
ODT. Cross-sectional structural (Fig. 3A) and veloc-
ity (Figs. 3B and 3C) images are obtained simultane-
ously in a single two-dimensional scan. The presence
of vessel-like circular features can be observed in the
structural image (Fig. 3A). Strong attenuation of the
light backscattered from locations deep in the skin indi-
cates a high degree of optical scattering. The dynamic
range of the measured backscattered light in Fig. 3A is
,45 dB. Velocity images of blood flow moving in op-
posite directions, as determined by the sign of the f
D
[Eq. (1)], are shown in Figs. 3B and 3C. Blood f low
in two small veins (Fig. 3B) and an artery (Fig. 3C)
is clearly identified. To demonstrate the versatility of
ODT, we also recorded en-face structural (Fig. 4A) and
velocity (Fig. 4B) images by scanning with the target
beam focused 200 mm below the skin surface. Blood
flow in a branching vessel is clearly evident. These
results demonstrate that ODT can be used for nonin-
vasive imaging of in vivo blood flow velocity in highly
scattering biological tissues.
The lateral and the axial spatial resolutions of
our ODT system are limited by the beam spot size
and the coherence length of the light source to 5
Fig. 2. ODT images of in vivo blood f low in a CAM vein.
A, color-coded structural image. B, color-coded velocity
image. C, velocity profile along a horizontal cross section
passing through the center of the vein. Open circles, ex-
perimental data; solid curve, fit to a parabolic function.
Fig. 3. ODT images of in vivo blood flow in rodent skin.
A, color-coded tomographic structural image. B, color-
coded velocity image of venous blood flow (into the page).
C, color-coded velocity image of arterial blood flow (out
of the page). Structural and velocity images are obtained
simultaneously from a single two-dimensional scan.
July 15, 1997 / Vol. 22, No. 14 / OPTICS LETTERS 1121
Fig. 4. A, en-face structural and B, velocity images of
blood flow in a vessel 200 mm below the skin surface.
and 13 mm, respectively.
6
Higher axial resolution can
be achieved if a source with a broader spectral emission
is used. Velocity resolution in our current system is
,100 mmys. This resoluton depends on the data-
acquisition time at each pixel and the angle between
the flow direction and the target beam. Resolution
can be improved if a smaller angle and (or) a longer
acquistion time is used. With our current scanning
speed of 800 mmys, data-acquisition time for an ODT
image of 1mm31mmwith 10-mm spatial resolution
is ,3min. The scanning speed is ultimately limited
by the data-acquisition time at each pixel, which affects
the detection sensitivity and the velocity resolution. If
we assume that detection of the Doppler shift requires
that one sample the interference-fringe intensity over
at least one oscillation cycle, pixel-acquisition time
varies inversely with f
D
. For example, resolving a
1-kHz Doppler shift, which corresponds to a velocity of
1000 mmys if the angle between the probe beam and
the flow direction is 70
±
, means that the minimum
data-acquisition time at each pixel is ,1ms. When
time variation of the interference-fringe intensity data
corresponding to each pixel is acquired sequentially,
a 100 3 100 pixel ODT image can be acquired in
10 s. However, in vivo blood f low at user-specified
locations can be probed with a response time of ,1ms
for a single pixel. The imaging speed in our current
ODT system is limited because temporal interference-
fringe intensity data are acquired serially. Shorter
image acquisition time (, 1s) is possible if other
scanning techniques are implemented, for example, if
the time variation of fringe intensity data at multiple
pixels is recorded in parallel.
ODT has great potential for use in the clinical
management of patients who can benefit from mi-
crovascular monitoring. The probing depth of ODT in
tissue is similar to that of optical coherence tomog-
raphy and is of the order of 1–2 mm, which covers
the epidermal and the dermal ranges. Information
provided by ODT could be used to monitor perfusion
and viability before, during, and after reconstructive
procedures, determine the efficacy of pharmacologi-
cal intervention for failing surgical skin f laps or re-
plants,
13
image microcirculation during sepsis, assess
burn depth,
14
diagnose atherosclerotic disease, and in-
vestigate the mechanisms of photodynamic therapy for
cancer treatment.
15
Although clinical applications of
ODT remain untested, we have used this technique
to monitor changes in blood flow dynamics and ves-
sel structure following photodynamic therapy.
16
ODT
could also be applied to other nonmedical areas in
which rapid noninvasive imaging of turbulent or lami-
nar f low is required. Given the noninvasive nature
of the measurement, exceptional spatial resolution,
simple hardware requirements, and relatively compact
size, ODT is a promising technique for both basic re-
search and clinical medicine.
We thank S. Kimel, D. J. Smithies, and T. Lindmo for
helpful discussions. This project is supported by re-
search grants awarded from the Biomedical Research
Technology Program and the Institute of Arthritis and
Musculoskeletal and Skin Diseases (1R29-AR41638-
01A1 and 1R01-AR42437-01A1) at the National Insti-
tutes of Health, the Whitaker Foundation (21025), and
the Dermatology Foundation. Institute support from
the U.S. Department of Energy (DE-FG03-91ER61-
227), the National Institutes of Health (RR-01192), and
the Beckman Laser Institute Endowment is also grate-
fully acknowledged.
References
1. Y. Yamada, M. Matsumura, S. Kyo, and R. Omoto, Am.
J. Cardiol. 75, 161 (1995).
2. J. V. Chapman, in Noninvasive Evaluation of Hemody-
namics in Congenital Heart Disease, J. V. Chapman
and G. R. Sutherland, eds. (Kluwer, Dordrecht, The
Netherlands, 1990), Chap. 2, pp. 57.
3. W. J. Manning, W. Li, and R. R. Edelman, N. Engl. J.
Med. 328, 828 (1993).
4. R. F. Bonner and R. Nossal, in Laser-Doppler Blood
Flowmetry, A. P. Shepherd and P. A. Oberg, eds.
(Kluwer, Dordrecht, The Netherlands, 1990), Chap. 2,
pp. 17.
5. D. Huang, E. A. Swanson, C. P. Lin, J. S. Schuman,
W. G. Stinson, W. Chang, M. R. Hee, T. Flotte, K.
Gregory, C. A. Puliafito, and J. G. Fujimoto, Science
254, 1178 (1991).
6. A. F. Fercher, J. Biomed. Opt. 1, 157 (1996).
7. J. A. Izatt, M. R. Hee, G. M. Owen, E. A. Swanson, and
J. G. Fujimoto, Opt. Lett. 19, 590 (1994).
8. Z. Chen, T. E. Milner, D. Dave, and J. S. Nelson, Opt.
Lett. 22, 64 (1997).
9. X. J. Wang, T. E. Milner, Z. Chen, and J. S. Nelson,
Appl. Opt. 36, 144 (1997).
10. X. J. Wang, T. E. Milner, and J. S. Nelson, Opt. Lett.
20, 1337 (1995).
11. J. G. Fujimoto, M. E. Brezinski, G. J. Tearney, S. A.
Poppart, B. Bouma, M. R. Hee, J. F. Southern, and
E. A. Swanson, Nat. Med. 1, 970 (1995).
12. S. Kimel, L. O. Svaasand, M. Hammer-Wilson, M. J.
Schell, T. E. Milner, J. S. Nelson, and M. W. Berns, J.
Invest. Dermatol. 103, 693 (1994).
13. P. C. Neligan, Microsurgery 14, 162 (1993).
14. Z. B. M. Niazi, T. J. H. Essex, R. Papini, D. Scott, N. R.
Mclean, and M. J. M. Black, Burns 19, 485 (1993).
15. M. Korbelik and G. Krosl, Br. J. Cancer 70, 604 (1994).
16. Z. Chen, T. E. Milner, S. Srinivas, and J. S. Nel-
son, ‘‘Optical Doppler tomography: in vivo blood flow
dynamics following pharmacologic intervention and
photodynamic therapy,’’ Photochem. Photobiol. (to be
published).
