Polarization sensitive optical coherence tomography of
melanin provides tissue inherent contrast based on
depolarization
The MIT Faculty has made this article openly available. Please share
how this access benefits you. Your story matters.
Citation
Baumann, Bernhard et al. “Polarization sensitive optical
coherence tomography of melanin provides tissue inherent
contrast based on depolarization.” Optical Coherence
Tomography and Coherence Domain Optical Methods in
Biomedicine XIV. Ed. Joseph A. Izatt, James G. Fujimoto, &
Valery V. Tuchin. San Francisco, California, USA: SPIE, 2010.
Vol.7554, 75541M-6. ©2010 SPIE--The International Society for
Optical Engineering.
As Published
http://dx.doi.org/10.1117/12.843226
Publisher
Society of Photo-optical Instrumentation Engineers
Version
Final published version
Accessed
Thu May 26 19:14:45 EDT 2016
Citable Link
http://hdl.handle.net/1721.1/57571
Terms of Use
Article is made available in accordance with the publisher's policy
and may be subject to US copyright law. Please refer to the
publisher's site for terms of use.
Detailed Terms
Polarization sensitive optical coherence tomography of melanin
provides tissue inherent contrast based on depolarization
Bernhard Baumann1,‡, Stefan O. Baumann2, Thomas Konegger3, Michael Pircher1, Erich Götzinger1,
Harald Sattmann1, Marco Litschauer2, Christoph K. Hitzenberger1
1
Center for Biomedical Engineering and Physics, Medical University of Vienna
2
Institute of Materials Chemistry, Vienna University of Technology
3
Institute of Chemical Technologies and Analytics, Vienna University of Technology
‡
Currently with Research Labs of Electronics, Massachusetts Institute of Technology and
New England Eye Center, Tufts Medical Center
ABSTRACT
Polarization sensitive optical coherence tomography (PS-OCT) was used to investigate the polarization properties of
melanin. Measurements in samples with varying melanin concentrations revealed polarization scrambling, i.e.
depolarization. The results indicate that the depolarizing appearance of pigmented structures like, for instance, the retinal
pigment epithelium (RPE) is likely to be caused by the melanin granules contained in these cells.
Keywords: Optical coherence tomography, polarization sensitive devices, imaging, melanin, retinal pigment epithelium,
depolarization, polarization scrambling
1. INTRODUCTION
Polarization sensitive (PS) optical coherence tomography (OCT) provides depth-resolved imaging of light polarizing
properties of transparent and translucent samples with micrometer scale resolution.1 Other than standard, intensity-based
OCT, which provides image contrast based on the amount of light backscattered from different sites inside the sample,
PS-OCT opens an additional contrast channel: By illuminating the sample with light in a well-defined polarization state
and by using a polarization sensitive detection unit, not only reflectivity but also the polarization state of light can be
assessed three-dimensionally. From the detected polarization state, phase retardation, optic axis orientation as well as
Stokes vectors can be calculated.1-4
In the human eye, PS-OCT imaging allows to distinguish structures with birefringent (e.g., cornea, retinal nerve fiber
layer), polarization preserving (e.g., stroma of the iris, photoreceptor layer), and depolarizing (e.g., pigment epithelia of
iris and retina) properties.3,5-7 It was shown that the unique depolarizing characteristics of the retinal pigment epithelium
(RPE) can be used for segmenting that layer.8 In diseased retinas, this enables to assess retinal disorders – such as fluid
volumes in exudative age-related macular degeneration (AMD) and the areas of atrophic zones in dry AMD – in a
quantitative way.9
Recently, we investigated the depolarizing character of the RPE in the macula region of healthy volunteers.10 The RPE’s
depolarization appeared most pronounced close to the center of the fovea and decreased in the periphery (Fig. 1).
However, the origin of the polarization scrambling was still unclear and could be caused by the outer morphology of the
RPE cells or by their structural composition. In addition to regular cell organelles like mitochondria or cell nuclei, the
human RPE cells contain different pigments in granular shape: melanin, lipofuscin, and composites of both.11 While the
lipofuscin concentration has a minimum in the fovea and increases in the periphery of the posterior pole, the
concentration of melanin shows the inverse characteristic with a peak concentration in the fovea (Fig. 2).12-14 Since the
Optical Coherence Tomography and Coherence Domain Optical Methods in Biomedicine XIV, edited by
Joseph A. Izatt, James G. Fujimoto, Valery V. Tuchin, Proc. of SPIE Vol. 7554, 75541M · © 2010 SPIE
CCC code: 1605-7422/10/$18 · doi: 10.1117/12.843226
Proc. of SPIE Vol. 7554 75541M-1
Downloaded from SPIE Digital Library on 29 Jul 2010 to 18.51.1.125. Terms of Use: http://spiedl.org/terms
macular distrribution of meelanin correspoonds to the deppolarization paattern measuredd with PS-OCT
T,10 one couldd suggest
that melaninn could be thee origin of thee RPE’s polariization scrambbling property.. In order to verify
v
this hyppothesis,
measurementts in phantoms are necessary..
Fig. 1. Maculaar depolarization
n pattern. (A) Foor the left eyes of
o 10 healthy volunteers, the minnimum degree of
o polarization uniformity
u
(DOPU) valuee was plotted fo
or each depth profile.
p
En face maps showing depolarization were
w
generated by plotting these values
DOPUmin for each
e
transverse position of the scanned
s
macularr regions. A maap showing the average
a
macularr DOPUmin distriibution of
the 10 eyes iss shown in (A). A decrease off the RPE’s depolarizing characcter in the peripphery can be obbserved. The poolarization
scrambling chharacter is most pronounced closse to the center of the fovea, which
w
manifests in the lowest DOPUmin values there.
t
(B)
Horizontal DO
OPUmin profiles extracted
e
from thhe macular depoolarization mapss are shown alonng with the meann profile (black line).
l
The
grey band indiicates the range of ±1 standard deviation arounnd the mean DO
OPUmin value. Likke in the en face map of (A), also
a in the
horizontal profiles a minimum
m of DOPUmin, i..e. most pronounnced depolarizattion, is visible inn the center of thhe fovea. A moree detailed
discussion of the
t depolarizatio
on patterns can be
b found in ref. [10]. (color onlinne)
Fig. 2. Distribuution of RPE pig
gments after ref. [13]. While lipoofuscin shows a local minimum in the central foovea, the melaninn
concentration has a maximum there. (color online)
Proc. of SPIE Vol. 7554 75541M-2
Downloaded from SPIE Digital Library on 29 Jul 2010 to 18.51.1.125. Terms of Use: http://spiedl.org/terms
2.
MATERIALS AND METHODS
Here, we present the results of PS-OCT measurements in melanin models. From raw eumelanin (Sigma Aldrich),
ellipsoid melanin particles were produced by dissolving and re-precipitating.15 A narrow size distribution of the melanin
particles with a maximum at ~530 nm was observed by dynamic light scattering (DLS) measurements (which
corresponds well to the size of melanin granules in the human RPE cells reported in literature).11 Scanning electron
micrographs of the powder revealed particles in the respective sizes and shapes (cf. Fig. 3). Using distilled water,
solutions with melanin concentrations Cmel = Vmel:VH2O in the range from 1:1 to 1:100 were prepared. The solutions were
treated with ultrasound for 30 s in order to prevent from and to break up particle agglomerates.
PS-OCT imaging was performed using our recently reported clinical PS-OCT instrument.16 In brief, we used a system
design based on a free-space Michelson interferometer with additional polarization optics. A broadband
superluminescent diode (center wavelength: 839 nm, FWHM bandwidth: 58 nm) was employed as a light source.
Circularly polarized light was used to illuminate the sample. At the interferometer exit, the OCT signal was split into two
orthogonal polarization components, which were detected by two identical spectrometer units. Each readout of the
spectrometer linescan cameras yielded the spectral data necessary to compute not only depth-resolved reflectivity, but
also phase retardation, optic axis orientation and Stokes vectors for one depth profile.
Using this PS-OCT system, 3D data sets (64×1024×1024 voxels, 12×12×3.3 mm³) of melanin solutions were recorded.
The liquid solutions were imaged inside a cylindrical glass container. Evaluation of a PS-OCT data set recorded in the
empty container confirmed the polarization preserving characteristics of the container material.
3. RESULTS AND DISCUSSION
Fig. 4 shows results of PS-OCT measurements of solutions with melanin concentrations of 1:3 and 1:60. Reflectivity Bscan images are shown on the left hand side (Figs. 4(A) and (B)). In order to evaluate polarization scrambling, degree of
polarization uniformity (DOPU) B-scan images were computed. For that purpose, the Stokes vector elements (I, Q, U, V)
were calculated from the amplitudes and phases of the two polarization channels’ OCT signals for each image pixel. The
Stokes vector elements were then normalized and averaged in a sliding evaluation window. DOPU(z) was then computed
2
2
2
as DOPU ( z ) = Qmean
( z ) + U mean
( z ) + Vmean
( z ) .8 For polarization preserving and birefringent structures, DOPU values
close to 1 will be observed, whereas polarization scrambling will result in lower DOPU values.
In Figs. 4(C) and (D), DOPU B-scans are shown for the concentrations of 1:3 and 1:60. Clearly, a polarization
scrambling characteristic can be observed, which manifests in low DOPU values for both concentrations. Also,
polarization scrambling appears more pronounced in the solution with the higher melanin concentration (Fig. 4(C)),
where the DOPU images are mostly bluish whereas the mostly green color in Fig. 4(D) indicates less polarization
scrambling for the lower concentration.
The polarization scrambling characteristic of melanin was investigated for concentrations from Cmel = 1:1 to Cmel = 1:100
by evaluating the distribution of DOPU values. The average DOPU value in the respective solutions is plotted in Fig. 5
as a function of melanin concentration. The polarization scrambling character is decreasing as the concentration of
melanin granules is decreasing.
The depolarizing characteristics observed in the melanin samples are consistent with the polarization scrambling
character of pigmented ocular structures. Therefore, the polarization scrambling appearance of the RPE (and that of the
iris pigment epithelium) is likely to be caused by the melanin granules.
4. CONCLUSION
In this study, we investigated the polarization properties of melanin samples. Ovoid melanin particles were produced and
solutions of concentrations in the range from 1:1 to 1:100 were prepared. PS-OCT data sets were recorded in these
Proc. of SPIE Vol. 7554 75541M-3
Downloaded from SPIE Digital Library on 29 Jul 2010 to 18.51.1.125. Terms of Use: http://spiedl.org/terms
solutions. Inn order to inveestigate the poolarization scrrambling charaacteristics of the acquired PS-OCT
P
data sets, an
evaluation off DOPU valuess was performeed. Depolarizattion (manifestiing in low DOP
PU values) waas more pronouunced for
higher concentrations of meelanin, and deccreased for low
wer melanin conncentrations. The
T polarizatioon scrambling character
c
of the melannin solutions was
w in analogyy to that of piggmented oculaar structures liike the RPE, which
w
contain melanin
granules of similar
s
size and
d shape. Thereffore, it is likelyy that the depoolarizing appeaarance of the pigment epithellia of iris
and retina bee caused by the melanin granuules contained within.
Fig. 3. Melaninn samples. (A) Size
S distribution as measured usiing DLS. (B) Meelanin solution. (C) – (F) Scanniing electron miccrographs
of melanin sam
mple granules at magnifications between 10,000× and 40,000×. (color online)
Proc. of SPIE Vol. 7554 75541M-4
Downloaded from SPIE Digital Library on 29 Jul 2010 to 18.51.1.125. Terms of Use: http://spiedl.org/terms
Fig. 4. PS-OCT
T B-scan images of melanin sam
mples. (A) – (B) Reflectivity imaages for melaninn concentrations of 1:3 and 1:60,
respectively. (C) – (D) DOPU images showingg increasing averrage DOPU for decreasing melaanin concentratioon. (color online)
Fig. 5. DOPU vs. melanin con
ncentration. For melanin
m
solutionns ranging from 1:1 to 1:100, avverage DOPU vaalues measured within
w
the
melanin solutiions are plotted. Clearly, a trendd from low to high DOPU valuees can be observved when the conncentration is deecreasing.
(color online)
Proc. of SPIE Vol. 7554 75541M-5
Downloaded from SPIE Digital Library on 29 Jul 2010 to 18.51.1.125. Terms of Use: http://spiedl.org/terms
ACKNOWLEDGEMENTS
The authors want to thank Christoph Wölfl at the Medical University of Vienna as well as Martin Wurm and David
Stifter at RECENDT GmbH for their excellent technical support. Financial support by the Austrian Science Fund (FWF
grants P19624-B02 and P19751-N20) is gratefully acknowledged.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
M. R. Hee, D. Huang, E. A. Swanson, J. G. Fujimoto, J. Opt. Soc. Am. B 9, 903-908 (1992).
J. F. de Boer, T. E. Milner, J. Stuart Nelson, Opt. Letters 24, 300-302 (1999).
B. Cense, T. C. Chen, B. Hyle Park, M. C. Pierce, J. F. de Boer, Invest. Ophthalmol. Vis. Sci. 45, 2606-2612 (2004).
C. K. Hitzenberger, E. Götzinger, M. Sticker, M. Pircher, A. F. Fercher, Opt.Express 9, 780-790 (2001).
M. Pircher, E. Götzinger, O. Findl, S. Michels, W. Geitzenauer, C. Leydolt, U. Schmidt-Erfurth, C. K. Hitzenberger,
Invest. Ophthalmol. Vis. Sci. 47, 5487-5494 (2006).
M. Miura, M. Yamanari, T. Iwasaki, A. E. Elsner, S. Makita, T. Yatagai, Y. Yasuno, Invest. Ophthalmol. Vis. Sci.
49, 2661-2667 (2008).
S. Michels, M. Pircher, W. Geitzenauer, C. Simader, E. Götzinger, O. Findl, U. Schmidt-Erfurth, C. K.
Hitzenberger, Br. J. Ophthalmol. 92, 204 - 209 (2008).
E. Götzinger, M. Pircher, W. Geitzenauer, C. Ahlers, B. Baumann, S. Michels, U. Schmidt-Erfurth, C. K.
Hitzenberger, Opt. Express 16, 16410-16422 (2008).
B. Baumann, M. Pircher, E. Götzinger, H. Sattmann, J. Jungwirth, C. Schütze, C. Ahlers, W. Geitzenauer, U.
Schmidt-Erfurth, C. K. Hitzenberger, Proc. SPIE 7372, 737207 (2009).
B. Baumann, M. Pircher, E. Götzinger, C. K. Hitzenberger, J. Biophoton. 2, 426 – 434 (2009).
K. M. Zinn, J. V. Benjamin-Henkind in :“The Retinal Pigment Epithelium”, edited by K. M. Zinn and M. F. Marmor
(Harvard University Press, Cambridge, Mass., 1979), chap. 1.
L. Feeney-Burns, W. S. Hilderbrand, S. Eldridge, Invest. Ophthalm. Vis. Sci. 25, 195-200 (1984).
J. J. Weiter, F. C. Delori, G. L. Wing, K. A. Fitch, Invest. Ophthalm. Vis. Sci. 27, 145-152 (1986).
C. N. Keilhauer, F. C. Delori, Invest. Ophthalm. Vis. Sci. 47, 3556-3564 (2006).
P. Giacomoni, L. Marrot, M. Mellul, A. Colette, PCT Int. Appl. WO 9425531 (1994).
B. Baumann, M. Pircher, E. Götzinger, H. Sattmann, M. Wurm, D. Stifter, C. Schütze, C. Ahlers, W. Geitzenauer,
U. Schmidt-Erfurth, C. K. Hitzenberger, Proc. SPIE 7163, 71630N (2009).
E. Götzinger, M. Pircher, C. K. Hitzenberger, Opt. Express 13, 10217-10229 (2005).
Proc. of SPIE Vol. 7554 75541M-6
Downloaded from SPIE Digital Library on 29 Jul 2010 to 18.51.1.125. Terms of Use: http://spiedl.org/terms