Assessment and Correction of Imaging Artifacts in
Skin Imaging using Fibre-based Optical Coherence
Yih Miin Liew1, Robert A. McLaughlin1, Andrea Curatolo1, and David D. Sampson1,2
1Optical+Biomedical Engineering Laboratory, School of Electrical, Electronic and Computer Engineering, M018
The University of Western Australia, 35 Stirling Hwy, Crawley 6009, Australia
2Centre for Microscopy, Characterisation & Analysis,
The University of Western Australia, 35 Stirling Hwy, Crawley 6009, Australia
Abstract Summary (35 words)
This study reports two types of imaging artifacts in 3D optical
coherence tomography images of human skin – intensity deficits and
geometrical distortion, and describes their reduction through the
application of refractive index matching media.
Keywords- Optical coherence tomography, skin imaging,
artifacts, index matching
Optical coherence tomography (OCT) is a high resolution,
optical in vivo imaging modality used in medical diagnostics
and research imaging. The working principle of OCT is
analogous to that of ultrasound ; however, instead of
acoustic waves, OCT employs broadband near infrared light to
probe biological tissue. The broadband light is delivered to the
tissue using a fibre-optic probe. Backscattered light from the
tissue is interfered with a reference beam with known optical
path length to provide precise optical gating of the tissue
microstructures. To date, a spatial resolution ranging from 2 to
20 microns is attainable with OCT.
Fibre-based OCT has been applied to the field of
dermatology for the assessment of skin diseases , tumours
, UV-induced skin alterations , skin characteristics , as
well as the efficacy of the laser treatment of port wine stain
OCT, in common with other imaging modalities, suffers
from imaging artifacts. Two dominant artifacts observed in
skin imaging are intensity and geometrical artifacts. Intensity
artifacts appear as multiple vertical streaks of artificially low
backscatter directly below the pits on the skin surface, as
shown in Fig. 1, causing the remapping of the surface
topography within the volume. This artifact masks the
subsurface structures and causes the appearance of remapped
extraneous features when viewing OCT en face images (i.e.,
images in the plane perpendicular to the light beam).
Geometrical artifacts are caused by changes in the optical
path length due to different proportions of tissue and air in
adjacent sections of the scan, and results in the artificial
displacement and shape distortion of subsurface structures.
OCT images of wrinkled skin, such as those on the arms, legs,
back of hand, eyelids and joints are commonly contaminated
with these artifacts. This artifact is illustrated in Fig. 4(A)
where a phantom sitting on a glass plate with a different
refractive index from air appears erroneously thick.
These artifacts were present in the OCT images of skin in
previous papers [2, 5, 7-9], but to our knowledge have not been
explicitly reported. This study addresses these two artifacts in
OCT, outlining possible causes and exploring their reduction
using two common refractive index-matching media – glycerol
and ultrasound gel.
MATERIALS AND METHODS
A. OCT system
All OCT images in this study were acquired using a swept-
source frequency-domain OCT system (OCS1300SS, Thorlabs,
Figure 1. (A) A volumetric skin OCT image shown in 3 orthogonal planes
(cs: cross section). (B) Cross-sectional image (cs 1) with intensity artifacts
(arrows). (C)Surface render of the volumetric image. (D) En face image (cs 2)
with intensity artifacts.
USA). This system (Fig. 2) contains a broadband, high-speed
frequency swept laser source and a fibre-based (SMF-28)
Michelson interferometer with balanced detection. The light
source has a centre wavelength of 1325 nm and an average
light output power of 10 mW. The system’s axial and
transverse resolution is 12 µm and 15 µm, respectively, in air.
The handheld OCT probe contains the focusing optics, and can
be easily positioned to scan a range of anatomical locations. A
glass cover slip (thickness 0.17 mm) was secured to the
handheld probe so as to be positioned immediately adjacent to
B. Phantom Study
A range of skin-mimicking silicone phantoms were
fabricated and consisted of an imprint of a human inner
forearm skin surface, with scattering provided by either TiO2 or
Al2O3 particles (Sigma Aldrich, USA). 3D scans of the
phantom were acquired with three different setups: a bare
phantom exposed to air; the phantom coated with a layer of
100% glycerol; and the phantom coated with a layer of
ultrasound gel. The glass cover slip affixed to the OCT probe
was used to flatten the refractive index matching media
(glycerol or ultrasound gel).
The refractive index of the phantom material (prepared
separately as a thin layer on a glass slide), glycerol and
ultrasound gel was measured according to the method proposed
in . The result is shown in Table 1.
C. Human Subject Study
A similar location on the skin of the inner forearm of a
human volunteer was imaged. Three scans were acquired: bare
skin; with a coating of 100% glycerol; and with a coating of
ultrasound gel. A time gap of 3 hours was left between scans
with glycerol and ultrasound gel to allow the skin to recover
from any possible swelling caused by the medium.
OCT images were acquired with the same imaging protocol
as used for the phantom study.
RESULTS AND DISCUSSION
A. Phantom Study
The OCT images of the skin-mimicking phantom are
shown in Figs. 3 and 4. Intensity and geometrical artifacts
were predominantly found in the image acquired without the
use of index matching medium.
Intensity artifacts characteristically emerge at different
depths under the pits of the surface corrugations and they
broaden with depth. We believe these artifacts are caused by
the strong scattering of off-normal incident light on the skin of
higher refractive index than air (of approximately 1.43). The
corrugated patterning of the skin was observed to be important
in the appearance of these artifacts.
The application of glycerol and ultrasound gel reduced the
intensity artifacts, delaying their onset to greater depths as
illustrated in Fig. 3. The en face (at approximately 1.5 mm
(optical distance) beneath the phantom surface) and cross-
sectional images of the phantom indicate that the application of
100% glycerol resulted in higher reduction in the intensity
artifact as compared to the ultrasound gel.
THE REFRACTIVE INDEX OF SILICONE PHANTOM (TIO2), 100%
GLYCEROL AND ULTRASOUND GEL
Medium Refractive Index, n
Ultrasound gel 1.34
As OCT displays depth of the image in terms of optical
path length, the image of the subsurface structures is artificially
Figure 3. Left: En face (at depth of ≈ 1.5mm (optical distance) from surface);
and Right: cross-sectional images of the skin-mimicking phantom using TiO2
scatterers. (A) Bare phantom exposed to air. (B) Phantom coated with a 100%
glycerol layer. (C) Phantom coated with a layer of ultrasound gel.
Figure 2. Fibre-based swept source OCT system with handheld probe for
in vivo imaging.
Figure 4. Geometrical artifact compensation in a phantom using Al2O3
scatterers, sitting on a glass plate using index matching media, imaged from
above. (A) Bare phantom exposed to air (Arrow: Discontinuity in the
appearance of the glass slide due to light traveling through different amounts
of phantom and air). (B) Phantom coated with a 100% glycerol layer. (C)
Phantom coated with a layer of ultrasound gel.
displaced and distorted when light travels through different
amount of tissue and air. This distortion can potentially result
in discontinuities or deformations in the shape of structures
such as blood vessels. Such artifacts were observed beneath the
pits of the skin surface corrugations.
Fig. 4(A) demonstrates the geometrical artifact in OCT
imaging, in which the thickness of a phantom sitting on top of
a glass plate and the surface location of the glass plate appear
distorted (arrow). This distortion was due to the uneven
thickness of the phantom, which can be compensated using
index matching medium and a glass cover slip. This is shown
in Figs. 4(B) and (C).
B. Human Subject Study
Similar results were obtained from in vivo images of human
skin. A reduction in intensity and geometrical artifacts was
observed using both index-matching media.
Fig. 5 shows the en face images of the inner forearm skin at
approximately 500 µm (optical distance) below the skin
surface. Glycerol and ultrasound gel were shown to reduce the
intensity artifacts in skin by reducing extraneous lines
mimicking surface topography in the en face images. In
addition, it was observed that glycerol gave better optical
clearing  of the skin than ultrasound gel, allowing better
visualization of tissue in deeper layers. This is due to the close
matching of its refractive index to tissue.
Fig. 6(A) shows an artifact-contaminated image of the inner
forearm skin with blood vessels, and the corresponding images
using index-matching media. Both intensity and geometrical
artifacts were observed to be reduced with the use of index
matching media (Figs. 6(B,C)). Note the artificial discontinuity
(bend) in the blood vessel (labeled with an arrow). The
geometrical distortions of the blood vessels due to the sudden
change in the thickness of the epidermal layer (and, thus,
optical path length of light) were compensated using both
index matching media.
Despite the reduction in image artifacts, there are
complications in the use of glycerol and ultrasound gel. This
includes trapping of air bubbles within the refractive index
matching medium which can cause the introduction of other
artifacts, including shadowing by the air bubbles. Careful
sample preparation is required to minimize the occurrence of
such air bubbles.
Artifacts are detrimental to the correct interpretation and
quantification of subsurface structures of the skin using OCT.
We have shown that a refractive index-matching medium can
be used to reduce intensity and geometrical artifacts in OCT
images of skin. Glycerol and ultrasound gels were assessed
using both phantom and in vivo image acquisitions. They are
both effective in reducing geometrical artifacts. Glycerol was
shown to give a better reduction in the intensity artifacts and
allows better optical clearing of the sample as compared to
T. Gambichler, G. Moussa, M. Sand, D. Sand, P. Altmeyer, and K.
Hoffmann, "Applications of optical coherence tomography in
dermatology," J. Dermatol. Sci., vol. 40, pp. 85-94, 2005.
J. Welzel, M. Bruhns, and H. H. Wolff, "Optical coherence tomography
in contact dermatitis and psoriasis," Arch. Dermatol. Res., vol. 295, pp.
F. G. Bechara, T. Gambichler, M. Stucker, A. Orlikov, S. Rotterdam, P.
Altmeyer, and K. Hoffmann, "Histomorphologic correlation with routine
histology and optical coherence tomography," Skin Res. Technol., vol.
10, pp. 169-173, 2004.
T. Gambichler, S. Boms, M. Stüker, G. Moussa, A. Kreuter, M. Sand, D.
Sand, P. Altmeyer, and K. Hoffmann, "Acute skin alterations following
ultraviolet radiation investigated by optical coherence tomography and
histology," Arch. Dermatol. Res., vol. 297, pp. 218-225, 2005.
T. Gambichler, R. Matip, G. Moussa, P. Altmeyer, and K. Hoffmann,
"In vivo data of epidermal thickness evaluated by optical coherence
tomography: Effects of age, gender, skin type, and anatomic site," J.
Dermatol. Sci., vol. 44, pp. 145-152, Dec 2006.
J. S. Nelson, K. M. Kelly, Y. Zhao, and Z. Chen, "Imaging blood flow in
human port-wine stain in situ and in real time using optical doppler
tomography," Arch. Dermatol., vol. 137, pp. 741-744, 2001.
J. Welzel, C. Reinhardt, E. Lankenau, C. Winter, and H. H. Wolff,
"Changes in function and morphology of normal human skin: evaluation
using optical coherence tomography," Br. J. Dermatol., vol. 150, pp.
Figure 5. En face images of the inner forearm skin at approximately 500 µm
from the skin surface. (A) bare skin (B) with a coating of 100% glycerol;
and (C) with a coating of ultrasound gel. (Note: All 3 images were
approximately co-located. Arrows: artifacts caused by trapped air bubbles in
the index matching medium.)
Figure 6. Cross-sectional images of the inner forearm skin showing the
geometrical artifacts (solid arrows) and intensity artifacts (dashed arrows) in
scans of the skin through: air (A); glycerol (B); and ultrasound gel (C).
J. Welzel, "Optical coherence tomography in dermatology: a review,"
Skin Res. Technol., vol. 7, pp. 1-9, 2001.
M. C. Pierce, J. Strasswimmer, B. H. Park, B. Cense, and J. F. de Boer,
"Birefringence measurements in human skin using polarization-sensitive
optical coherence tomography," J. Biomed. Opt., vol. 9, p. 287, 2004.
 G. J. Tearney, M. E. Brezinski, J. F. Southern, B. E. Bouma, M. R. Hee,
and J. G. Fujimoto, "Determination of the refractive index of highly
scattering human tissue by optical coherence tomography," Opt. Lett,
vol. 20, pp. 2258-2260, 1995.
 Z. Mao, D. Zhu, Y. Hu, X. Wen, and Z. Han, "Influence of alcohols on
the optical clearing effect of skin in vitro," J. Biomed. Opt., vol. 13, pp.