UC Irvine
UC Irvine Previously Published Works
Title
Ultrahigh-resolution optical coherence tomography by broadband continuum generation
from a photonic crystal fiber.
Permalink
https://escholarship.org/uc/item/3ww18454
Journal
Optics letters, 28(3)
ISSN
0146-9592
Authors
Wang, Yimin
Zhao, Yonghua
Nelson, JS
et al.
Publication Date
2003-02-01
DOI
10.1364/ol.28.000182
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
182 OPTICS LETTERS / Vol. 28, No. 3 / February 1, 2003
Ultrahigh-resolution optical coherence tomography
by broadband continuum
generation from a photonic crystal fiber
Yimin Wang, Yonghua Zhao, J. S. Nelson, and Zhongping Chen
Beckman Laser Institute and Department of Biomedical Engineering, University of California, Irvine, Irvine, California 92612
Robert S. Windeler
Bell Laboratories, Lucent Technologies, Murray Hill, New Jersey 70974
Received September 20, 2002
We have developed an ultrahigh-resolution optical coherence tomographic system in which broadband contin-
uum generation from a photonic crystal fiber is used to produce high longitudinal resolution. Longitudinal
resolution of
1.3-mm has been achieved in a biological tissue by use of continuum light from 800 to 1400 nm.
The system employed a dynamic-focusing tracking method to maintain high lateral resolution over a large
imaging depth. Subcellular imaging is demonstrated. © 2003 Optical Society of America
OCIS code: 170.4500.
Optical coherence tomography (OCT) can be used
for in vivo high-resolution cross-sectional imaging
of biological tissues.
1–3
Longitudinal resolution,
governed by source coherence length, is inversely
proportional to the light source’s bandwidth. Super-
luminescent diodes (SLDs) are often used for OCT
imaging and typically have 10
15-mm longitudinal
resolution.
4
However, the limited optical bandwidth
of SLDs precludes imaging on a cellular level. In
addition, achieving faster imaging speed with a high
signal-to-noise ratio requires more than the milliwatt-
level power that is typically available from SLD
sources. A Kerr-lens mode-locked Ti
:
sapphire laser,
optimized for a short coherence length, has achieved
sub-2-mm longitudinal resolution.
5
Although this
source provided cellular resolution imaging, com-
parative studies have shown that the 0.8-mm center
wavelength of the Ti
:
sapphire laser is not optimal
for deep penetration into highly scattering biological
tissues. Ultrahigh-resolution OCT imaging in the
spectral region from 1.1 to 1.5 mm requires extremely
broad bandwidths because of the l
2
兾Dl dependence of
the longitudinal resolution. However, this spectral
region is of particular interest for OCT because it
penetrates deeply into biological tissue and permits
spectrally resolved imaging of water absorption bands.
In vivo, OCT imaging with 5.1-mm longitudinal reso-
lution at 1300 nm was demonstrated with a Kerr-lens
mode-locked solid-state laser.
6
However, the im-
provement in resolution is constrained by the limited
output bandwidth of the laser.
Highly-nonlinear air –silica microstructure fibers
and photonic crystal fibers (PCFs) can generate an
extremely broadband continuum light spectrum from
the visible to the near infrared by use of low-energy
femtosecond pulses.
7
This spectrum is achieved
because of the waveguide dispersion characteristics of
the fibers, which shift the zero-dispersion wavelength
to shorter wavelengths, and the small core diameters,
which provide tight mode conf inement. An OCT
system with a longitudinal resolution of 2.5 mm with
a PCF source at a wavelength of 1.3 mm has been
reported.
8
However, the system has a limited depth
range because no dynamic focus tracking is imple-
mented. In this Letter we describe the application of
a PCF source for ultrahigh-resolution OCT imaging.
In our experiment, 1.3-mm longitudinal resolution
was achieved in tissue with a center wavelength
of 1.1 mm. To our knowledge, this is the highest
resolution reported for this wavelength range. The
system also incorporated a dynamic focus tracking
method that significantly increased the depth range
of ultrahigh-resolution imaging. Subcellular imaging
over a large depth range has been demonstrated.
Fiber-based systems have been widely used in
OCT. However, several limitations become apparent
when a fiber-based OCT system is used for high-
resolution imaging. First, it is difficult to construct a
2 3 2 coupler with a bandwidth of more than 200 nm.
Second, because of chromatic aberration, a lens sys-
tem with fiber collimators or focusing elements will
limit the eff icient bandwidth. Third, the fiber would
be multimode if the short-wavelength limit of the
spectrum were shorter than the cutoff wavelength of
the fiber. The multiple modes will introduce ghost
lines into the OCT image. Finally, polarization mode
dispersion in a single-mode fiber may also limit lon-
gitudinal resolution. To overcome these limitations,
we designed and constructed an open-air ultrahigh-
resolution OCT system, as shown in Fig. 1. To
achieve high lateral resolution it is necessary to
use an objective lens with a high numerical aper-
ture to obtain a small focusing size. However, a
small focusing size also means a small confocal
parameter, which is connected with the longitu-
dinal imaging range. To get deep longitudinal
imaging while maintaining constant high lateral reso-
lution, several methods were developed to overcome
that problem. One example is the use of C-mode
scanning to reconstruct tomographic images, but this
increases image acquisition time. In our system,
an objective lens and a rectangular prism are both
0146-9592/03/030182-03$15.00/0 © 2003 Optical Society of America
February 1, 2003 / Vol. 28, No. 3 / OPTICS LETTERS 183
Fig. 1. Schematic of the ultrahigh-resolution OCT sys-
tem: M
3
, wideband cube beam splitter; M
1
,M
2
,M
4
,M
5
,
silver mirrors; F
1
, long-pass f ilter; L
1
, coupling lens; L
2
,
collimated lens; L
3
, objective lens; P, rectangular prism;
D, InGaAs detector; F
2
, is a neutral-density f ilter; PCF,
photonic crystal fiber; FS, fused-silica prisms.
mounted upon a voice-coil translation stage to track dy-
namic focusing during longitudinal scanning.
9
This
makes it possible to increase imaging depth with
constant high lateral resolution. The baseplate is
mounted upon a stage for lateral scanning.
The interferometer’s light source is broadband
continuum generation from a PCF, as shown in
Fig. 1. The pump source for the PCF was a Kerr-lens
mode-locked Ti
:
sapphire laser. The total output
laser power was greater than 700 mW, with a pulse
duration of 110 fs and a repetition rate of 76 MHz.
The laser output wavelength was 780 nm. The laser
beam was then coupled into the PCF after a Faraday
isolator to prevent interference of backref lected light
with the mode locking and the
l兾2 wave plate, as
shown in Fig. 1. The focal length of the coupling lens
was 6.5 mm. Nitrogen gas was slowly blown onto
the coupling part to purge the f iber tip and prevent
damage. A l兾2 wave plate after the Faraday isolator
was used to adjust the polarization state of the light
input to the fiber to optimize the spectrum. The
spectrum of light was broadened as it propagated
through the fiber because of self-phase modulation
and Raman scattering. The continuum output of
light was collimated by a 4.5-mm focal-length lens,
L
2
. Continuum light generation from 400 to 1400 nm
was observed from the fiber. The output spectrum
after passing through an 800-nm long-pass filter is
shown in Fig. 2(a). The spectrum ranged from 800 to
1430 nm. The total output power from the PCF could
be as high as 100 mW, and the remaining power was
approximately 50 mW after a long-pass filter.
To achieve ultranarrow-field correlation it is impor-
tant to balance the group-velocity dispersion (GVD) of
the two arms of the interferometer. For a light source
with a Gaussian spectrum, the width of the field auto-
correlation function increases according to
10
s
t
苷 s
t0
Ω
1 1
∑
d
2
f共v兲
dv
2
∏
2
s
v
4
æ
1兾2
, (1)
where s
t
is the 1兾e
1兾2
half-width of the autocorre-
lation function, s
t0
is the zero-dispersion width of
the autocorrelation function, and s
w
is the 1兾e
1兾2
half-width of the spectrum. The equation shows that
the larger the source bandwidth, the more sensitive
to GVD the measurement is. For this reason, we
chose a pair of prisms made from fused silica to
balance the system GVD and used two SLD sources,
with wavelengths of 650 and 950 nm. When GVD
in the system was not compensated for, the fringes
of 650 and 950 nm were separated from each other.
Through adjustment of the prism, the distance be-
tween these fringes could be eliminated. When
the GVD was balanced, the two fringes perfectly
overlapped. Thus, with this method, the GVD in
the system could be accurately balanced. In the
experiment the ultrahigh-resolution OCT system
was optimized to support 1.8-mm longitudinal reso-
lution in free space at a center wavelength of 1.1 mm.
Figure 2(b) shows the interference fringe with a silver
mirror as the sample. The envelope is shown in
Fig. 2(d), and it can be seen that the longitudinal
resolution is 1.8 mm. Considering the refractive
index of tissue, the corresponding resolution is
1.3 mm in tissue. Because the spectrum from the
PCF is not perfectly Gaussian shaped, sidelobes
in the interferometric autocorrelation are visible.
The detected optical spectrum is shown in Fig. 2(c),
which was calculated by Fourier transformation of
the interferometric signal shown in Fig. 2(b). The
detected spectral bandwidth is reduced to 372 nm at a
center wavelength of 1.1 mm. This reduction of the
bandwidth occurs in the short-wavelength part of the
spectrum and is caused by the low sensitivity of our
detector at wavelengths shorter than 0.9 mm.
The lateral resolution of the system was 4 mm,
which was determined by the achromatic lens. The
improvement in axial resolution was demonstrated in
the following experiment: A plastic film of 21-mm
thickness was chosen as the sample. Two glass plates
were used to clip the film. Index-matching oil was
put onto the upper and lower sides of the film to reduce
ref lection. The power of the light to the system was
50 mW. The OCT image is shown in Fig. 3. The im-
age size was 0.75 mm 3 0.15 mm at ⬃1.8 mm 3 4 mm
(longitudinal 3 lateral) resolution. Note that the
Fig. 2. (a) Spectrum of the light source from the crystal
fiber after a long-pass f ilter, (b) interference signal at the
photodetector, (c) spectrum calculated by Fourier transfor-
mation, (d) interference envelope of the signal shown in (b).
184 OPTICS LETTERS / Vol. 28, No. 3 / February 1, 2003
Fig. 3. OCT image with a 21-mm-thick plastic film
sample.
Fig. 4. (a) OCT image of a tadpole for which the
dynamic focusing technique was used. Image size,
0.5 mm 3 1.0 mm. (b) OCT image of the tadpole without
dynamic focusing. Image size, 0.5 mm 3 1.0 mm.
structure inside the film can be observed. In particu-
lar, the gap between the glass plate and upper surface
of the f ilm can be clearly seen. Compared with the
21-mm film thickness, the calibrated thickness of the
gap at A is 2 mm. Thus, with broadband continuum
light from a PCF source, OCT axial resolution has
been significantly improved.
The feasibility of in vivo ultrahigh-resolution
imaging was demonstrated with an animal model,
a Xenopus tadpole. Figure 4(a) shows an in vivo
OCT image of a 35-day-old anesthetized specimen.
The tadpole was placed in tank water in a dish to
prevent dehydration and to provide index matching.
The incident power to the system was 50 mW. The
voice coil stage was scanned at a speed of 4 mm
兾s
at a repetition rate of 1 Hz. Figure 4(a) shows an
area of 0.5 mm 3 1.0 mm imaged at 1.3 mm 3 4 mm
(longitudinal 3 lateral) resolution. Membranes and
nuclei can be clearly seen, and single-cell structure
can be identified in the image.
To demonstrate the advantage of dynamic focus-
ing for improved imaging depth while maintaining
high lateral resolution, another experiment was
performed. Here a normal interference structure
without dynamic focusing was constructed. The OCT
image of a 21-day-old specimen is shown in Fig. 4(b).
The resolution in Fig. 4(b) is 1.3 mm 3 4 mm
(longitudinal 3 lateral). Single cells can be seen, but
the imaging depth is greatly reduced compared with
that of the image obtained by the dynamic-focusing
technique. Because of the high lateral resolution,
the confocal parameter of our objective lens was only
33.8 mm in air, which resulted in image degradation
outside the focused zone. Thus, dynamic focusing has
been employed successfully in ultrahigh-resolution
OCT imaging.
In summary, we have demonstrated an ultrahigh-
resolution optical coherence tomography system that
uses broadband-continuum generation from a photonic
crystal fiber as the light source. We achieved a longi-
tudinal resolution of 1.8 mm in air 共⬃1.3 mm in tissue)
at a center wavelength of 1.1 mm. The broad band-
width of the light source permits both high-resolution
and spectroscopic OCT imaging in wavelength ranges
that were previously not accessible. The system is ca-
pable of dynamic focusing compensation to maintain
constant lateral resolution for large imaging depth.
The authors are grateful to L. Newman and D. M.
Gardiner for supplying and preparing the tadpoles.
Y. Wang thanks Zifu Wang and Hongwu Ren for
their help with the experiment. This research was
supported by research grants awarded by the National
Science Foundation (BES-86924) and the National
Institutes of Health (EB-00255, EB-00293, and
RR-01192). Institutional support from the U.S. Air
Force Office of Scientific Research (F49620-00-1-0371)
and from the Beckman Laser Institute Endowment
is also gratefully acknowledged. Z. Chen’s e-mail
address is zchen@bli.uci.edu.
References
1. D. Huang, E. 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).
2. J. M. Schmitt, S. L. Lee, and K. M. Yung, Opt. Com-
mun. 142, 203 (1997).
3. J. G. Fujimoto, M. E. Brezinski, G. J. Tearney, S. A.
Boppart, B. E. Bouma, M. R. Hee, J. F. Southern, and
E. A. Swanson, Nature Med. 1, 970 (1995).
4. R. C. Youngquist, S. Carr, and D. E. N. Davies, Opt.
Lett. 12, 158 (1987).
5. W. Drexler, U. Morgner, F. X. Kärtner, C. Pitris, S. A.
Boppart, X. D. Li, E. P. Ippen, and J. G. Fujimoto, Opt.
Lett. 24, 1221 (1999).
6. S. A. Boppart, B. E. Bouma, C. Pitris, J. F. Southern,
M. E. Brezinski, and J. G. Fujimoto, Nature Med. 4,
861 (1998).
7. J. K. Ranka, R. S. Windeler, and A. J. Stentz, Opt. Lett.
25, 25 (2000).
8. I. Hartl, X. D. Li, C. Chudoba, R. K. Ghanta, T. H. Ko,
and J. G. Fujimoto, Opt. Lett. 26, 608 (2001).
9. Z. Chen, T. E. Milner, D. Dave, and J. S. Nelson, Opt.
Lett. 22, 64 (1997).
10. B. L. Danielson and C. Y. Boisrobert, Appl. Opt. 30,
2975 (1991).
