Download high-resolution imaging in nontransparent tissue

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 5, NO. 4, JULY/AUGUST 1999
1185
Optical Coherence Tomography: High-Resolution
Imaging in Nontransparent Tissue
Mark E. Brezinski and James G. Fujimoto
Abstract—Optical coherence tomography (OCT) is a method of
high-resolution imaging originally developed for the transparent
tissue of the eye. Recently, the technology has been advanced
toward the difficult challenge of imaging in nontransparent tissue.
In this paper, three topics are addressed. First, the principles
behind OCT imaging are discussed. Second, the difficulties associated with OCT imaging in nontransparent tissue are outlined.
Finally, the feasibility of OCT for medical imaging will be
discussed. Specifically, OCT demonstrates its greatest potential
in situations where conventional biopsy is either dangerous or
ineffective.
Index Terms— Dysplasia, imaging, intravascular ultrasound,
light, optical coherence tomography, plaque rupture.
I. INTRODUCTION
O
PTICAL coherence tomography (OCT) was originally
developed for imaging in transparent tissue [1]. Recent
advances have resulted in imaging in nontransparent tissue
with an imaging resolution approaching that of histopathology [2]–[4]. This has led to the suggestion it will replace
conventional biopsy in many circumstances. As this powerful
technology moves into clinical trials, issues arise as to where
it should be applied, both with respect to patient safety and
controlling health care costs. The appropriate clinical scenarios
for application of this technology need to be identified since,
in many circumstances, it is difficult to justify replacing a
gold standard like excisional biopsy. For example, in a patient
suspected of having malignant melanoma, a life threatening
skin tumor, excisional biopsy is a very effective technique
for making the diagnosis. It is debatable whether replacing
the reliable technique of excision biopsy with an imaging
modality is prudent. However, there are many situations where
conventional biopsy is not effective or is associated with a
high complication rate. It is in these scenarios where OCT
demonstrates its greatest potential. In this manuscript, data
will be presented supporting a role for OCT in three general
areas of clinical medicine where the application of excisional
biopsy is of limited value. In addition, the principles behind
OCT imaging in nontransparent tissue will be reviewed.
Manuscript received January 5, 1999; revised April 26, 1999. M. E.
Brezinski was supported in part by the National Institutes of Health under
Contract NIH-RO1-AR44812-01 and Contract NIH-1-R29-HL55686-01A1,
and in part by the Whitaker Foundation under Contract 96-0205. J. G.
Fujimoto was supported in part by the Contract NIH-9-RO1-CA75289-01 and
Contract NIH-9-RO1-EY11289-10, and in part by the Medical Free Electron
Laser Program, Office of Naval Research under Contract N00014-94-1-0717.
M. E. Brezinski is with Deparment of Medicine, Harvard Medical School,
Massachusetts General Hospital, Boston, MA 02114 USA.
J. G. Fujimoto is with the Department of Electrical Engineering and Computer Science, Research Laboratory of Electronics, Massachusetts Institute of
Technology, Cambridge, MA 02139 USA.
Publisher Item Identifier S 1077-260X(99)07507-3.
A. Optical Coherence Tomography
The principles behind OCT have been previously described
[1]. OCT is analogous to ultrasound, measuring the intensity
of reflected infrared light rather than sound waves. Time gating
is employed so that the time for the light to be reflected
back, or echo delay time, is used to assess the intensity
of backreflection as a function of depth. Unlike ultrasound,
the echo delay time can not be measured electronically due
to the high speed associated with the propagation of light.
Therefore, a Michelson interferometer is used to perform low
coherence interferometry [5]–[7]. If the source illuminating
the interferometer generates light with a broad bandwidth, the
autocorrelation function representing the interference is then
proportional to
where
is the intensity at the detector,
is the product
is
of the reflections off the sample and mirror,
the real component of the Fourier transform of the power specis the center frequency of the source,
trum of the source,
is the phase delay [8]. The width of the spectrum and
and
the width of the autocorrelation function (coherence length)
are inversely related via the Fourier transform. Therefore, the
resolution increases (shorter coherence length) with increasing
source bandwidth. If the source has a Gaussian spectrum with
, and a
a full-width at half-maximum (FWHM) bandwidth,
or axial resolution is
center , then the coherence length
Through the use of sufficiently broad source bandwidths,
resolutions with OCT have been achieved in the range of 4
to 16 m, up to 25 higher than high-frequency ultrasound.
High-frequency ultrasound (30 MHz) is the current endoscopic
or catheter based procedure with the highest resolution [9].
The lateral or transverse resolution achieved with an OCT
imaging system is determined by the focused spot size in
analogy with conventional microscopy [10]. The transverse
resolution is
where is the spot size on the objective lens and
is its
focal length. High transverse resolution can be obtained by
using a large numerical aperture and focusing the beam to a
small spot size. In addition, the transverse resolution is also
related to the depth of focus or confocal parameter (b). The
.
confocal parameter is two times the Raleigh range
1077–260X/99$10.00  1999 IEEE
1186
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 5, NO. 4, JULY/AUGUST 1999
Its relationship to transverse resolution is described by the
formula
In addition to its high resolution, advantages of OCT for
medical imaging include its broad dynamic range, rapid data
acquisition rate, small inexpensive catheter/endoscope design,
and compact portable structure. In order to obtain backreflecting data as a function of depth, the optical pathlength in the
reference arm is changed. The rate at which the pathlength
is changed generally determine the acquisition rate of the
system. The frame rate for OCT systems are currently at four
to eight frames per second [11]. All frame rates described in
this text will assume an image size of 256 by 512 pixels.
Initially, the pathlength in the reference arm was changed via
the use of a moving mirror or galvanometer [1]. However, this
embodiment required approximately 40 seconds to perform
an image of nontransparent tissue [2]. A system similar to
this is still in use for imaging the transparent tissue of the
eye. Second generation systems changed the pathlength in
the reference arm via fiber stretching with a peizoelectric
crystal [12], [13]. Limitations of this embodiment included
polarization mode dispersion, hysteresis, crystal breakdown,
and voltage requirements unacceptably high for routine clinical
use [13]. A recent embodiment showing considerable promise
induces a variable optical group delay in the reference arm
through the introduction of a grating-based phase control delay
line [8]. Light from the reference arm is directed onto a grating,
which results in a Fourier transform of the spectrum. The
dispersed light is directed onto a mirror. By tilting the mirror,
a linear wavelength dependent phase ramp is introduced. A
linear phase ramp in the frequency domain results in an optical
group delay in the time domain. When the light is redirected
onto the grating, an inverse Fourier transform occurs. By
varying the angle of the mirror, the optical group delay can
be varied. In addition to high data acquisition rates (4–8
frames/second), this system has two additional advantages
over previous embodiments. The optical group delay can be
varied separately from the phase delay, and the group velocity
dispersion can be varied without the introduction of a separate
prism [11].
Imaging deep within nontransparent tissue is possible because of the high sensitivity and dynamic range of OCT. The
system functions analogously to heterodyne optical detection
and achieves nearly quantum limited performance. OCT has
been designed near the shot noise limit by choosing a Doppler
frequency (induced by the motion of the mirror or fiber
noise and a proper
stretching) to avoid low frequency
transimpedance amplifier resistance/reference arm voltage to
overcome thermal noise [14]. For shot noise detection, under
the assumption of linearity of electronics and infinite dynamic
range of the digitization electronics can be expressed as
SNR
NEB
where
is the number of electrons per unit time
generated by the detector due to returning light and NEB
band pass filter bandwidth [14].
An additional advantage of OCT, besides high data acquisition rates and high resolution, is the fact that it is fiber optically
based, making possible the development of small catheters and
endoscopes. The small size is important for clinical viability.
The current OCT catheter/endoscopes are 1 mm in diameter
[11]. A significant advantage of OCT catheter/endoscopes is
that unlike an ultrasound catheter, they contain no transducer
within their frame, making them relatively inexpensive. They
consist primarily of relatively simple components, a single
mode optical fiber, grins lens, light directing prism, and
speedometer cable. Since the diameter of early prototypes was
dependent primarily on the size of the available speedometer
cables, it is likely that commercial devices will be substantially
smaller.
B. Imaging in Nontransparent Tissue
OCT was originally developed to image the transparent
tissue of the eye at unprecedented resolution [1]. It has
been used clinically to evaluate a wide range of retinalmacular diseases [15]–[17]. Recently, the technology has been
advanced to image in nontransparent tissue, where penetration
of light is very limited [2]–[4].
Light penetration in tissue is dependent on the absorption
and scattering properties of the tissue [18]. Tissue absorption
of light in the visible region is high and results predominately
from electronic transitions. The amount of absorption in the
near infrared is much lower and is due primarily to transitions
in molecular vibrational states, with some contribution from
changes in rotational states. Scattering is a relatively complex
phenomena which has its origin in refractive index mismatches
[19]. The intensity and angular dependence of scattering is
determined by the size, position, and shape of the scatterer
relative to the wavelength of incident light in addition to the
index mismatch.
OCT imaging in nontransparent tissue requires optimizing
light penetration through the tissue [2]. Improved light penetration is achieved by performing imaging with light having an
incident wavelength near 1300 nm. At this wavelength, light
scattering is low relative to scattering of light in the visible
region. Absorption is low because this wavelength is too long
to result in large amounts of electron transitions but is too
short for inducing extensive vibrational transitions in water,
which has a maximum near 1450 nm. The advantage of OCT
imaging at 1300 nm is illustrated in Fig. 1 of human epiglottis.
The top image was performed at 850 nm, the middle image
at 1300 nm, and the bottom is the histology. Penetration has
been dramatically increased at 1300 nm.
C. Application of OCT
As stated above, OCT demonstrates its greatest potential
in situations where conventional biopsy is currently of limited
clinical value. We have identified at least three general clinical
scenerios for the application of OCT. The first is situations
where conventional biopsy is difficult or impossible to perform
because of the hazards involved. This would include biopsies
of the brain, coronary artery, and cartilage surface of joints.
The second situation where biopsy is ineffective occurs in
BREZINSKI AND FIJIMOTO: OPTICAL COHERENCE TOMOGRAPHY
1187
Fig. 2. Schematic of the OCT system. The system, the heart of which is a
Michelson interferometer, is described within the text.
Fig. 1. Influence of median wavelength on penetration. Imaging was performed of a human epiglottis at two wavelengths (850 and 1300 nm) and
compared with histology. At 850 nm, most of the data is obtained at the
surface of the image. At 1300 nm, structural detail is identified deeper within
the tissue, with the underlying cartilage clearly identified. The bar in all images
represents 500 m unless indicated (reproduced with permission from [2, p.
1206]).
clinical scenarios associated with high “false negative” rates.
An example is the early diagnosis of cancers. Many malignancies, such as uterine and esophageal cancers, are screened
for in high risk patients, but are difficult to detect at curable
stages. A high false negative rate means that the diagnosis
was missed, even though the physician was looking for it. The
third situation where OCT demonstrates its greatest potential
is in guiding microsurgical procedures, such as the repair
of peripheral nerves and blood vessels. In this manuscript,
OCT imaging of tissue from these three scenarios will be
demonstrated, defending its feasibility for high resolution
diagnostic imaging.
II. METHODS
A. In Vitro Experiments
Human tissue was obtained either postmortem or postsurgical resection. The samples were stored in 0.9% saline
with 0.1% sodium azide at 0 C. Imaging was performed
on segments smaller than 10 cm by 10 cm. The samples
were selected based on their macroscopic appearance. The
samples were placed on a translation stage, which permits
the acquisition of multiple depth scans resulting in a two-
dimensional cross-sectional image. The position of the invisible infrared light beam on the sample was monitored with
a visible light guiding beam. The peripheral areas of the
imaged sections were marked with microinjections of dye to
permit comparison of the OCT-images with histologic images.
The histologic processing of the sample included fixing in
10% formalin for 12 h, dehydration, and paraffin embedding. Different microstructures were identified by staining
with hematoxylin/eosin (H/E) and trichrome blue. Stained
histologic sections were compared with OCT images to allow
correlation.
Images with 15- m axial resolution were obtained with a
superluminescent diode, with a median wavelength of 1300 nm
and a bandwidth of 50 nm. The median wavelength is defined
as the wavelength of maximum intensity within the Gaussian
spectrum. The resolution was determined experimentally by
measuring the axial point spread function using a mirror. The
lateral resolution was 30 m, with a corresponding confocal
parameter of 520 m. The measured signal-to-noise ratio
(SNR) was 109 dB, using a power of 160 W at the sample.
The SNR was measured by measuring the maximum signal
when the optical beam was reflected from a high reflecting
mirror divided by the background noise level of the instrument.
Acquisition rates varied from 5 to 40 s during in vitro imaging,
depending on the image pixel size.
B. In Vivo Experiment
A self-phase modulated KLM Cr :Forsterite laser was
developed and used as the low coherence source for the in
vivo system [11]. The laser was set to produce an output power
of 30 mW with a Gaussian FWHM spectral bandwidth of 75
nm, centered at 1280 nm. These parameters corresponded to
a measured free space axial resolution of 10 m and an SNR
for the laser of 110 dB. The sample arm power was 10 mW.
This high-speed scanning optical delay line, as described
above, is shown in Fig. 2 and allows acquisition of 2000 axial
scans per second. The overall SNR of the OCT system coupled
into the catheter-endoscope was approximately 106 dB. The
1188
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 5, NO. 4, JULY/AUGUST 1999
Fig. 4. Vulnerable plaque. High-resolution imaging is needed in cardiology
to identify the small plaque which lead to most heart attacks, often referred to
as vulnerable plaque. In this image, a high risk plaque is seen within the wall
of the aorta. The dark region within the wall is lipid and the arrow identifies
a thin intimal cap vulnerable to rupture (reproduced with permission from [2,
p. 1206]).
Fig. 3. OCT imaging catheter/endoscope. The 2.9 French (1 mm) OCT
imaging catheter is shown. Unlike an ultrasound catheter, it contains no
transducer within its frame and consists of relatively simple fiber optics,
making it inexpensive.
image acquisition time was approximately 250 ms. The total
optical path length delay produced was approximately 2.8 mm.
A 2.9 French (1 mm in diameter) OCT imaging
catheter/endoscope (next to a dime) was used for in vivo
imaging in the rabbit, as shown in Fig. 3. The catheter consists
of relatively simple components, a single mode optical fiber,
Gradient Index (GRIN) lens, prism, and speedometer cable.
The measured working distance of the catheter-endoscope was
approximately 3.0 mm from the central axis. The focussed
beam diameter was 40 m, corresponding to a confocal
parameter of 1.9 mm. The image size was 512 by 248 pixels.
Closed loop feedback electronics of the motor controller
unit enabled synchronization of the motor rotation with the
high-speed OCT frame acquisition. Data was recorded in both
super VHS and digital format.
Animal handling was performed in accordance with MIT
and MGH Committee on Animal Care guidelines, protocol. At
the age of 12 weeks, a normal New Zealand white rabbit was
anesthetized with ketamine (35 mg/kg), xylazine (5 mg/kg),
and glycopyrrolate (0.01 mg/kg), given intramuscularly. Maintenance of the anesthesia was accomplished with intravenous
ketamine (8 mg/kg) and xylazine (1 mg/kg), administered via
a marginal ear vein. During the procedure, local injections of
lidocaine were given when necessary.
For both gastrointestinal tract and cardiovascular system
imaging, the OCT imaging catheter/endoscope was introduced
through a seven French guiding catheter. The gastrointestinal
tract was entered via the oropharynx while the aorta was accessed through a midabdominal incision. Imaging in the aorta
was performed in the presence and absence of saline flushes
to assess the effects of blood on imaging. This procedure was
repeated three or four times, washing the transparent window
of the endoscope with each pass. After imaging, the rabbit
was sacrificed with an intra-arterial injection of 5-ml sodium
pentobarbital (65 mg/ml). The imaged regions were excised
and immersed in 10% formalin in preparation for routine
histologic processing.
Fig. 5. Comparison of OCT and High Frequency Ultrasound. High- frequency ultrasound is the current clinical technology with the highest resolution. Catheter based OCT was directly compared with high-frequency
ultrasound (30 MHz). The superior resolution with OCT is evident by the
sharp delineation of arterial layers (reproduced from Tearney et al., Circulation
4256, 1997).
III. RESULTS
AND
DISCUSSION
A. Biopsy Hazardous
Probably the most important tissue for OCT application,
which can not be biopsied because of the hazards involved, is
vascular tissue. The prevalent pathology of the vascular system
is atherosclerosis, the replacement of normal tissue with a
lipid and fibrous mixture called plaque, which is the cause of
most heart attacks or myocardial infarctions [20]. Myocardial
infarction is the leading cause of death in the industrialized
world [21]. It results from the abrupt loss of blood flow to
a region of the heart, resulting in death to that heart tissue.
Most heart attacks result when small plaques (rather than the
large ones seen with conventional techniques) rupture. When
they rupture, they release fat into the blood stream, clot forms,
and the vessel occludes [20]. The vulnerable region of these
plaques are less than 200 m in diameter, which is below the
detection limit of the high-frequency ultrasound (30 MHz),
the current clinical technology with the highest resolution [22].
Therefore, a true clinical need exists for an imaging technology
with sufficient resolution to identify these lesions prior to
rupture.
We have demonstrated the feasibility of OCT for the identification of high risk or vulnerable plaque that lead to heart
attacks [2]. In Fig. 4, a high-risk plaque is seen with a
vulnerable region less than 50 m in diameter (arrow). In
Fig. 5, OCT is directly compared with high frequency ultra-
BREZINSKI AND FIJIMOTO: OPTICAL COHERENCE TOMOGRAPHY
1189
Fig. 6. In vivo OCT vascular imaging. Imaging was performed of the rabbit
aorta in vivo using the OCT imaging catheter. Imaging was performed at 4
frames per second. A saline flush was required during imaging due to the high
scattering of blood (reproduced with permission from [24]).
sound (30 MHz) [18]. The superior delineation of structure
is clear. This is consistent with quantitative measurements,
which demonstrated an axial resolution of OCT at 16 1
m and 110 7 m for ultrasound [23]. In Fig. 6, in vivo,
intravascular imaging was performed in a rabbit at four frames
per second through an OCT catheter 1 mm in diameter. Saline
flushes are required for imaging the vascular system since
blood substantially reduces penetration. [24].
Besides the vascular system, OCT has demonstrated feasibility for high-resolution imaging of other vulnerable tissue.
This includes the central nervous system and the cartilage of
joints [25], [26].
Fig. 7. Imaging gastrointestinal malignancies. The top OCT image is of
normal esophagus, where the squamous epithelium has a homogeneous
pattern. In the middle image, adenomitousepithelium is seen with a regular
appearance of the glands. In the lower image, an early adenocarcinoma is seen
with the glands completely lost on the left and dilated/distorted on the right
(modified from [30, pp. 1800–1804]).
B. High False-Negative Rates
In general, a high false negative rate means that the diagnosis was missed, even though the physician was looking for it.
This is a particularly important problem when screening highrisk patients for cancers. An important example is esophageal
cancer occurring in patients with “heartburn” or esophageal
reflux [27]. Reflux is the chronic presence of stomach acid
in the esophagus. About 1 million Americans with reflux
will develop Barrett’s Esophagus, which is the change in the
surface of the esophagus from normal to “colon-like.” The
colon-like surface has a 40 risk of developing adenocarcinoma [27]. The current method for screening patients with
Barrett’s Esophagus is to do endoscopy on 18 month intervals
where during the procedure, random biopsies are performed
every centimeter in the hope of detecting a malignancy at
an early stage. This screening procedure is expensive and of
only limited success. A clinical need exists for an imaging
technology capable of identifying these cancers at early stages.
The ability of OCT to provide high speed imaging of the
esophagus near the resolution of histology makes it well suited
for the screening of Barrett’s Esophagus.
In Fig. 7, we see normal esophagus (top), colonic epithelium (middle), and an early cancer (bottom). It can be seen
Fig. 8. In vivo rabbit esophagus. OCT imaging was performed of the rabbit
esophagus at 8 frames/secondthrough a 2.9 French (1 mm) OCT imaging
catheter. Structural layerswithin the esophagus are sharply defined (reproduced
with permission from [11, pp. 2037–2039]).
that normal esophagus and colonic epithelium have distinct
appearances by OCT. Furthermore, in the early cancer, the
surface glands are lost on the left and have become dilated
and distorted on the right, suggesting feasibility for OCT in
the early diagnosis of malignancies.
In Fig. 8, an in vivo OCT image of the rabbit esophagus
is seen. This image was generated through a 1-mm OCT
endoscope at 4 frames/s. The surface layers of the esophagus
are sharply defined [11].
1190
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 5, NO. 4, JULY/AUGUST 1999
Fig. 9. Imaging the prostate. Guiding prostatic resection has important
implications in reducing the incidence of impotence and incontinence. In this
image, the prostate capsule (left) is sharply differentiated from the prostate
proper (right). A small vulnerable nerve is seen within the capsule (arrow). It
is when these vulnerable nerves are transected during prostatic resection that
impotence and incontinence result ([28, p. 157).
In addition to esophageal cancer, OCT has demonstrated
considerable promise in the early diagnosis of cancers of the
bladder, uterus, colon, cervix, and lung [28]–[31].
C. Guiding Microsurgical Procedures
The surgeon, by addressing pathology on smaller scales,
now effectively manages pathology which were previously
beyond the scope of traditional medical therapeutics [25].
Improved methods for assessing microstructure, in addition
to technical refinements in tissue manipulation, have been
the cornerstone of these advances in surgical interventions
[32]–[33]. However, the high complication rates associated
with many of these procedures are in large part due to limitations of current imaging modalities. A technology capable
of providing the surgeon with superior, real-time information
about tissue microstructure and yet has a low profile within
the operative field would meet many of these needs. The
feasibility of using optical coherence tomography to provide
high-resolution real-time imaging for surgical diagnostics and
guidance has been demonstrated [25].
There are two important areas for OCT guidance of surgical
procedures, the prevention of iatrogenic injury (which is injury
caused by the surgeon) and the guidance of microsurgical
repair. An important example of where the prevention of
iatrogenic injury is needed is in resection of the prostate [28].
The resection of the prostate, often referred to as a TURP, is
one of the most common surgical procedures performed in this
country. It is performed when an enlarged prostate obstructs
urine flow. Two of the most feared complications of this
procedure are impotence and incontinence, which results when
nerves in the capsule (and not the prostate body) are damaged
during the procedure [28]. Fig. 9 not only demonstrates the
ability of OCT to distinguish the prostate capsule from the
body (top arrows), but also identify the small vulnerable nerves
within the capsule (bottom arrows).
Another important example where surgical guidance is
important for preventing iatrogenic injury is the resection of
brain tumors, where small amounts of normal tissue injury can
have catastrophic consequences [25]. The two important areas
where OCT imaging is important in guiding repair are in nerve
Fig. 10. In vivo cellular level imaging. Through the use of a source with a
broader bandwidth, the resolution of OCT imaging was increased to 4 m.
This allowed true subcellular imaging within this living Xenopus organizm,
with not only cells seen, but true subcellular imaging [36].
and vessel reanastamosis, which have been covered in great
depth elsewhere [25], [35].
D. Future Work/Limitations
The current limitations of OCT are the penetration, resolution, acquisition rate, and lack of large-scale clinical trials.
The maximum penetration of OCT is currently between 2–3
mm, dependent on the tissue type examined. Although the use
of alternative median wavelengths (e.g., 1500–1750) may lead
to small improvements in penetration, there is currently no
evidence that penetration greater than 4 mm will be possible
in the near future. Therefore, applications which require significantly greater penetration, such as the staging of metastatsis,
will not likely be possible with OCT. The poor penetration
through blood, an important issue for intravascular imaging,
is being approached currently with saline flushing [24].
The axial resolution of current catheter/endoscope based
OCT systems is 10 m. Resolutions in the range of 4 m
have been achieved with noncatheter based systems through
the use of broad bandwidth solid-state lasers. In Fig. 10,
true subcellular level imaging is demonstrated in vivo of a
tadpole, where not just cells are seen but individual nuclei
can be identified [36]. However, to routinely implement this
for catheter/endoscope based imaging, two advances need
to occur. First, the mode-locked femtosecond lasers used to
generate Fig. 9 are too complex and expensive for routine
clinical use. Therefore, clinically viable sources need to be
developed with similar median wavelength, power, and bandwidth to those of the mode locked lasers. Second, during
imaging, transverse resolution needs to be similar to axial
resolution. To match the transverse resolution to an the axial
resolution below 10 m, a very short confocal parameter is
used which results in the focus falling off rapidly. To alleviate
this problem, the focus needs to be scanned in a manner
analogous to confocal microscopy, requiring more complex
catheter/endoscope designs than the current prototypes.
As stated, maximum data acquisition rates are currently
under 10 frames/second. With improved methods for rapidly
changing the optical group delay in the reference arm, video
rate is anticipated with future embodiments. Finally, clinically
trials are underway assessing the advantages and limitations of
OCT over a wide range of applications, with only pilot studies
currently in the literature [37]. The results of these trials will
be critical in defining directions for future basic research on
OCT technology.
BREZINSKI AND FIJIMOTO: OPTICAL COHERENCE TOMOGRAPHY
1191
[2]
[3]
[4]
[5]
[6]
[7]
Fig. 11. Polarization sensitive OCT imaging. In this figure, a section of
articular cartilage was imaged with incident light in three different polarization states. Unlike most tissue, imaging ofnormal cartilage changes with
polarization variations of the incident light. This is due to the highly organized
structure of the collagen within the cartilage, which is birefringent. Since
disorganization of the collagen is one of the earliest signs of osteoarthritis,
the loss of this birefringence can lead to early detection of osteoarthritis,
independent of measurements of cartilage width [26].
Work is underway which combine OCT with Doppler
velocimetry and measurement of birefringence properties [26],
[38]–[40]. This has the potential of giving OCT the ability to
make both structural and dynamic assessments. In Fig. 11, polarization sensitive OCT imaging is seen [26]. In this image, a
section of articular cartilage was imaged with incident light in
three different polarizations states. Unlike most tissue, imaging
of normal cartilage changes with polarization variations of the
incident light. This is due to the highly organized structure of
the collagen within the cartilage, which is birefringent. Since
disorganization of the collagen is one of the earliest signs of
osteoarthritis, the loss of this birefringence can lead to early
detection of osteoarthritis.
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
IV. CONCLUSION
OCT represents an attractive new technology for performing
high resolution imaging when conventional biopsy is not
effective. This includes situations where conventional biopsy
is hazardous, when high false negative rates are present, or in
the guidance of surgical procedures.
ACKNOWLEDGMENT
The work on OCT in nontransparent tissue represents the
effort of a large number of students, Post-Doctoral Fellows,
and collaborators. This includes Dr. M. Hee, Dr. G. Tearney,
Dr. B. Bouma, Dr. S. Boppart, Dr. J. Southern, Dr. J. Izatt,
Dr. D. Stamper, Dr. P. Pitwari, Dr. J. Herrmann, C. Pitris, and
Dr. N. Weissman. The authors also appreciate the technical
support of C. Jesser and the secretarial support of C. Kroft.
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
REFERENCES
[1] 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.
[27]
G. Fujimoto, “Optical coherence tomography,” Science, vol. 254, pp.
1178–1181, 1991.
M. E. Brezinski, G. J. Tearney, B. E. Bouma, J. A. Izatt, M. R. Hee,
E. A. Swanson, J. F. Southern, and J. G. Fujimoto, “Optical coherence
tomography for optical biopsy: Properties and demonstration of vascular
pathology,” Circulation, vol. 93, pp. 1206–13, 1996.
J. M. Schmitt, A. Knuttel, M. Yadlowsky, and M. A. Eckhaus, “Opticalcoherence tomography of a dense tissue: Statistics of attenuation and
backscattering,” Phys. Med. Biol., vol. 39, pp. 1705–1720, 1994.
J. M. Schmitt, M. Yadlowsky, and R. F. Bonner, “Subsurface imaging
of living skin with optical coherence microscopy,” Dermatol., vol. 191,
pp. 93–8, 1995.
K. Takada, I. Yokohama, K. Chida, and J. Noda, “New measurement
system for fault location in optical waveguide devices based on an
interferometric technique,” Appl. Opt., vol. 26, pp. 1603–1606, 1987.
R. C. Youngquist, S. Carr, and D. E. N. Davies, “Optical coherence
domain reflectometery: A new opitcal evaluation technique,” Opt. Lett.,
vol. 12, pp. 158–160, 1987.
A. F. Fercher, K. Mengedoht, and W. Werner, “Eye-length measurement
by interferometry with partially coherent light,” Opt. Lett., vol. 13, pp.
1867–1869, 1988.
N. Wiener, “Generalized harmonic analysis,” Acta Math., vol. 55, pp.
117–122, 1930.
P. J. Benkeser, A. L. Churchwell, C. Lee, and D. M. Aboucinaser,
“Resolution limitations in intravascular ultrasound imaging,” J. Amer.
Soc. Echocardiol., vol. 6, pp. 158–165, 1993.
T. R. Cole and G. S. Kino, Confocal Scanning Optical Microscopy and
Related Imaging Systems. San Diego, CA: Academic, 1990.
G. J. Tearney, M. E. Brezinski, B. E. Bouma, S. A. Boppart, C. Pitris, J.
F. Southern, and J. G. Fujimoto, “In vivo endoscopic optical biopsy with
optical coherence tomography,” Science, vol. 276, pp. 2037–2039, 1997.
A. Sergeev, V. Gelikonov, and A. Gelikonov, “High-spatial-resolution
optical-coherence tomography of human skin and mucous membranes,”
presented at the Conf. Lasers and Electro Optics’95, Anaheim, Ca, May
21–26, 1995.
G. J. Tearney, B. E. Bouma, S. A. Boppart, B. Golubovic, E. A.
Swanson, and J. G. Fujimoto, “Rapid acquisition of in vivo biological
images by use of optical coherence tomography,” Opt. Lett., vol. 21,
pp. 1408–1410, 1996.
P. R. Gray and R. G. Meyer, Analysis and Design of Integrated Circuits,
2nd ed. New York: Wiley, 1984.
M. R. Hee, J. A. Izatt, E. A. Swanson et al., “Optical coherence
tomography of the human retina,” Arch. Opthalmol., vol. 113, pp.
325–32, 1995.
C. A. Puliafito, M. R. Hee, C. P. Lin et al., “Imaging of macular diseases
with optical coherence tomography,” Ophthalmol., vol. 120, pp. 217–29,
1995.
C. A. Puliatifo, M. R. Hee, J. S. Schumann, and J. G. Fujimoto,
Optical Coherence Tomography of Ocular Diseases. Thorofare, NJ,
Slack, 1995.
P. Parsa, S. Jacques, and N. Nishioka, “Optical properties of rat liver
between 350 and 2200 nm,” Appl. Opt., vol. 28, pp. 2325–2330, 1989.
J. J. Bowman, T. B. Senior, and P. L. E. Uslenghi, Electromagnetic and
Acoustic Scattering by Simple Shapes. New York: Summa, 1987.
V. L. Fuster, L. Badimon, J. J. Badimon et al., “The pathogenesis of
coronary artery disease and the acute coronary syndromes,” New Eng.
J. Med., vol. 326, pp. 242–249, 1992.
1996 Statistics on Cardiovascular Disease and Stroke, Amer. Heart
Assoc., Dallas, TX.
H. M. Loree, R. D. Kamm, R. G. Stringfellow, and R. T. Lee,
“Effects of fibrous cap thickness on peak circumferential stress in model
atherosclerotic vessels,” Circ. Res., vol. 71, pp. 850–858, 1992.
M. E. Brezinski, G. J. Tearney, N. J. Weissman et al., “Assessing
atherosclerotic plaque morphology: Comparison of optical coherence tomography and high frequency ultrasound,” Heart, vol. 77, pp. 397–403,
1997.
J. G. Fujimoto, S. A. Boppart, G. J. Tearney, B. E. Bouma, C. Pitris,
and M. E. Brezinski, “High resolution in vivo intraarterial imaging with
optical coherence tomography,” Heart, vol. 82, pp. 128–133, 1999.
M. E. Brezinski, G. J. Tearney, S. A. Boppart, E. A. Swanson, J. F.
Southern, and J. G. Fujimoto, “Optical biopsy with optical coherence
tomography, feasibility for surgical diagnostics,” J. Surg. Res., vol. 71,
pp. 32–38, 1997.
J. Herrmann, C. Pitris, B. E. Bouma, S. A. Boppart, J. G. Fujimoto, and
M. E. Brezinski, “High resolution imaging of normal and osteoarthritic
cartilage with optical coherence tomography,” J. Rheumatol., vol. 26,
pp. 627–35, 1999.
R. W. Phillips and R. K. H. Wong, “Barrett’s esophagus: Natural history,
1192
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 5, NO. 4, JULY/AUGUST 1999
incidence, etiology, and complications,” Gastro. Clinics North Amer.,
vol. 20, pp. 791–815, 1991.
G. J. Tearney, M. E. Brezinski, J. F. Southern, B. E. Bouma, S. A.
Boppart, and J. G. Fujimoto, “Optical biopsy in human urologic tissue
using optical coherence tomography,” J. Urol., p. 157, 1997.
C. Pitris, A. K. Goodman, S. A. Boppart, J. J. Libus, J. G. Fujimoto,
and M. E. Brezinski, “High resolution imaging of cervical and uterine
malignancies using optical coherence tomography,” Obstect. and Gyn.,
vol. 93, pp. 135–139, 1999.
G. J. Tearney, M. E. Brezinski, J. F. Southern, B. E. Bouma, S. A.
Boppart, and J. G. Fujimoto, “Optical biopsy in human gastrointestinal
tissue using optical coherence tomography,” Amer. J. Gastro., vol. 92,
pp. 1800–1804, 1997.
C. Pitris, M. E. Brezinski, B. E. Bouma, G. J. Tearney, J. F. Southern,
and J. G. Fujimoto, “High resolution imaging of the upper respiratory
tract with optical coherence tomography,” Amer. J. Respiratory and
Critical Care Med., vol. 157, pp. 1640–1644, 1998.
J. H. Culbertson, R. P. Rand, and M. J. Jurkiewicz, “Advances in
microsurgery,” Adv. Surg., vol. 23, p. 57, 1990.
D. S. McManamny, “Comparison of microscope and loupe magnification: Assistance for the repair of median and ulnar nerves,” Br. J.
Plastic Surg., vol. 36, p. 367, 1983.
S. N. Roper, M. D. Moores, G. V. Gelikonov, F. I. Feldchstein, N.
M. Beach, M. A. King, V. M. Gelinkonov, A. M. Sergeev, and D. H.
Reitze, “In vivo detection of experimentally induced cortical dysgenesis
in the adult rat neocortex using optical coherence tomography,” J.
Neuroscience Methodol., vol. 80, pp. 91–98, 1998.
S. A. Boppart, B. E. Bouma, C. Pitris, G. J. Tearney, J. F. Southern, M.
E. Brezinski, and J. G. Fujimoto, “Three-dimensional optical coherence
tomography for microsurgical diagnostics,” Radiol., vol. 208, pp. 81–86,
1998.
S. A. Boppart, B. E. Bouma, C. Pitris, J. F. Southern, M. E. Brezinski,
and J. G. Fujimoto, “In vivo subcellular optical coherence tomography
imaging in Xenopus laevis: Implications for the early diagnosis of
neoplasms,” Nature-Medicine, vol. 4, pp. 861–865, 1998.
A. M. Sergeev, V. M. Gelikonov, G. V. Gelikonov, F. I. Feldchtein,
R. V. Kuranov, N. D. Gladkova, N. M. Shakhova, L. B. Snopova, A.
V. Shakhov, I. A. Kuznetzova, A. N. Denisenko, V. V. Pochinko, Y.
P. Chumakov, and O. S. Streltzova, “In vivo endoscopic OCT imaging
of precancer and cancer states of human mucosa,” Opt. Express, vol. 1,
pp. 432–440, 1997.
J. A. Izatt, M. D. Kulkarni, S. Yazdanfar, J. K. Barton, and A. J. Welch,
“In vivo bidirectional color doppler flow imaging of picoliter blood
volumes using optical coherence tomography,” Opt. Lett., vol. 22, pp.
1439–1441, 1997.
[39] J. F. de Boer, T. E. Milner, M. J. C. van Germert, and J. S. Nelson,
“Two dimensional birefringence imaging in biological tissue by polarization sensitive optical coherence tomography,” Opt. Lett., vol. 22, pp.
934–936, 1997.
[40] M. R. Hee, D. Huang, E. A. Swanson, and J. G. Fujimoto, “Polarization
sensitive low coherence reflectometer for birefringence characterization
and ranging,” J. Opt. Soc. Amer. B, vol. 9, pp. 903–908, 1992.
Mark E. Brezinski received the undergraduate degree was from King’s
College, the M.D. and Ph.D. degrees from Thomas Jefferson University.
He is a full-time basic scientist, an assistant professor at Harvard Medical
School, and staff member at Massachusetts General Hospital. He did his PostDoctoral research fellowships at Harvard University from 1988 to 1992. His
residency in internal medicine was at Brigham and Women’s Hospital and
his cardiology fellowship was at Massachusetts General Hospital. He has
authored and published 91 invited presentations, papers, and book chapters
and has seven patents issued or submitted in the fiber-optic imaging field.
Along with E. Swanson and J. Fujimoto, he is a co-founder of Coherent
Diagnostic Technologies, Concord, MA.
Dr. Brezinski has received 14 awards for basic science research including
the 1999 Presidential Early Career Award from President Clinton.
James G. Fujimoto received the undergraduate, M.S., and Ph.D. degrees from
the Massachusetts Institute of Technology, Cambridge, MA.
He is a Professor in the Department of Electrical Engineering and Computer
Science, Massachusetts Institute of Technology and Adjunct Professor of
Ophthalmology at Tufts University. He has received multiple awards including
the National Academy of Sciences Baker Award for Initiatives in Research.
He is the author of over 150 publications and two hundred abstracts. He is
the author of three books Optical Coherence Tomography of Ocular Diseases
(1995); Ultrafast Phenomena X (1996); and Advances in Optical Imaging
and Photon Migration (1996). He also holds seven patents and has five
additional patent applications pending. He has conducted extensive research in
the development and application of femtosecond laser technology and studies
of ultrafast phenomena, and is very active in laser medicine and surgery,
including the development of optical coherence tomography imaging. He
also co-founded Coherent Diagnostic Technology with E. Swanson and M.
Brezinski in 1998.