Open Access
Add to BibSonomyAbstract
Recently, we developed a phase resolved polarization sensitive OCT system based on transversal scanning. This system was now improved and adapted for retinal imaging in vivo. We accelerated the image acquisition speed by a factor of 10 and adapted the system for light sources emitting at 820nm. The improved instrument records 1000 transversal lines per second. Two different scanning modes enable either the acquisition of high resolution B-scan images containing 1600×500 pixels in 500ms or the recording of 3D data sets by C-scan mode imaging. This allows acquiring a 3D-data set containing 1000×100×100 pixels in 10 seconds. We present polarization sensitive B-scan images and to the best of our knowledge, the first 3D-data sets of retardation and fast axis orientation of fovea and optic nerve head region in vivo. The polarizing and birefringence properties of different retinal layers: retinal pigment epithelium, Henle’s fiber layer, and retinal nerve fiber layer are studied.
©2004 Optical Society of America
1. Introduction
Optical coherence tomography (OCT) represents a powerful imaging technique to obtain cross sectional images with a depth resolution at the order of a few micrometer [1–3]. The human retina represents the main application field of OCT and several reports have shown the diagnostic value of this technique [4–6]. Recently, fast imaging techniques have been introduced, which enable 3D and real time imaging [7–10]. However, most of these imaging techniques did not provide information on the birefringent properties of the sample. On the other hand, it is known from other optical polarization sensitive imaging techniques that some parts of the retina (e.g., retinal nerve fiber layer (RNFL) and Henle fibers) are birefringent [11, 12] and therefore change the polarization state of light. The polarizing properties can be used to obtain additional information on the retina. E.g. scanning laser polarimetry (SLP) measures the amount of retardation introduced by the RNFL to obtain information on thickness changes associated with glaucoma. However, this method cannot provide depth resolved information on the polarizing properties of the retina. Polarization sensitive low coherence interferometry (PS-LCI) [13] and polarization sensitive OCT (PS-OCT) combine the ability of high depth resolution imaging with the ability to measure birefringent properties of a sample [14, 15]. A variety of birefringent tissues (skin, cornea, retina) have been investigated with PS-OCT so far [16–21]. The latter reports [19, 21] provided quantitative depth resolved data (obtained by averaging in lateral dimension) of the retardation of the retinal nerve fiber layer close to the nerve head.
While conventional OCT is based on A-scans, i.e., the priority or fast scan is in z-direction perpendicular to the retinal surface, an alternative approach which is based on a transversal scanning of the sample can be found in references [22, 23]. In this technique the fast scan direction (x) is parallel to the retinal surface. For each depth position of the coherence gate a complete (x-y) image is recorded. First transversal polarization sensitive devices which measured the retardation were introduced, recently [24, 25]. A very fast variant of this technique based on frequency generation by acousto optic modulators was recently demonstrated [7]. This technique provided no information on the birefringent properties of the sample. A phase and polarization sensitive system based on this transversal scanning technique has been introduced recently and was demonstrated in imaging applications of the anterior segment [26].
In this paper we present an improved and adapted transversal phase resolved polarization sensitive OCT system and show 2D-images and 3D-data sets of intensity, retardation and fast axis orientation of regions of fovea and optic nerve head of human retina in vivo. These images and data sets are used to characterize the polarizing properties of several different retinal layers.
2. Method
The method is based on a transversal scanning of the retina. A more detailed description of the basic experimental setup and the operating principle of the technique can be found in reference [26]. The system consists of a Mach Zehnder interferometer which is illuminated by a vertically polarized short coherence light source. A quarter wave plate placed in the sample arm generates circular polarized light which is incident on the sample. For in vivo retinal imaging, the instrument was modified in several ways. To avoid the spectral region of strong light absorption of aqueous and vitreous humor, the 1310nm light source was replaced by an SLD operating at a center wavelength of 821nm (FWHM bandwidth Δλ=25nm). To obtain the imaging speed necessary for 3D imaging the carrier frequency which determines the maximum transversal scanning velocity, was shifted from 100kHz to 1MHz. Our system generates this carrier frequency as the frequency difference of two acousto optic modulators (AOM’s) that are traversed by the reference beam [26]. The transversal scanning device was adapted by introducing an additional lens between scanner and eye to generate a collimated beam illuminating the cornea of the eye. The pivot point of the x-y scanner is thereby imaged into the pupil plane of the eye. A glass rod of appropriate length was placed into the reference arm to compensate a dispersion mismatch between the reference arm and sample arm which is introduced by the ocular media and by the additional lens. At each exit of the Mach Zehnder interferometer the light was directed to a polarization sensitive detection unit. Light in the horizontal polarization state of each exit was detected by the first balanced detector (Femto HCA-S). Light in the vertical polarization state was detected by the second balanced detector. Each signal (corresponding to the horizontal and vertical polarization state) was amplified and high pass filtered with an analog low noise amplifier (Stanford Research Systems SR560). The whole interferometric signal was recorded with a data acquisition board (NI-PCI6110) at a sampling rate of 4Ms/sec. Each signal was band pass filtered and phase sensitive demodulated by software. From the measured data we can calculate the sample reflectivity R(z)[13]
where A1 denotes the amplitude of the horizontal channel, and A2 denotes the amplitude of the vertical channel, respectively. The (single pass) retardation δ is given by the quotient of the amplitudes [13]
The fast axis orientation θ is encoded in the phase difference ΔΦ=Φ2-Φ1 between the two channels [27]
The cumulative axis orientation of the sample can be measured in an unambiguous range of 180°. The power illuminating the eye was measured with 1.2mW which is safe according to ANSI standards for repeated exposures within measuring time [28]. The measuring time for 3D- measurements was 10sec with an average scanning spot speed on the retina of approximately 5 m/s. A normal retina of the left eye of a human volunteer was measured without pupil dilation.
The system can be operated either in a high resolution B-scan mode (x-z-image plane) or in a fast C-scan mode (en-face image plane). The modified system records 1000 transversal lines (scans in x direction) per second and enables a more than 10 times faster image acquisition than the previous system [26]. Note that in most OCT images the lateral dimension exceeds the depth dimension and therefore the number of A-scans which are necessary to build up an OCT image of a given size and resolution exceeds the number of transversal lines. (Imagine an image size of 5mm (lateral) x 1mm (depth) with an equal lateral and depth resolution of 10µm. The necessary sampling points (without over sampling) in depth are 100 pixels in contrast to 500 pixels in lateral direction. This means that an A-scan based system with 1000 A scans per second needs 500 A scans (~0.5 seconds) to record the image, whereas a transversal system with 1000 transversal lines per second needs only 100 transversal lines (~0.1 seconds)) Therefore the image acquisition of a transversal system recording 1000 transversal lines per second is faster than that of a standard OCT system recording 1000 A-scans per second. A B-scan image consisting of 1000(x)×500(z) pixels is recorded in 0.5 seconds. The acquisition of a 3D data set (1000(x)×100(y)×100(z) pixels) takes 10 seconds. The scanning angle of the beam can be varied up to 15 degree. A real time display (4 en-face images per second) enables a fast and simple aligning of the eye.
3. Results
Figure 1 shows images of intensity, (single pass) retardation and cumulative fast axis orientation of the fovea region of a human volunteer in vivo. Each image consists of 1600×500 (x-z) pixels and covers an area of ~5×1mm2. The backscattered intensity in Fig. 1(a) is displayed on a logarithmic scale, with white pixels representing low backscattering and black pixels representing highly backscattering areas. Although a light source yielding standard OCT resolution (~12µm depth resolution in air) was used, all layers which can be seen in ultra high resolution images [29] could be identified in the intensity image (c.f., Fig. 1(a)). Furthermore, in the central part of the fovea a separation of the layer which is labeled retinal pigment epithelium (RPE) into two layers is visible (c.f., Fig. 1(d)). This corresponds to results obtained with a dual beam technique [30] and to results obtained with an ultrahigh resolution Fourier domain OCT technique reported recently [31, 32] (note that the separation of these layers is larger than the coherence length of our source). An explanation or comparison with anatomical structures of this additional layer visible in OCT images has not yet been presented. The two layers might be identified as RPE and Bruchs membrane and will be named in the following as RPE/Bruchs-membrane (RPE/BM) complex. Although the influence of anterior segment birefringence was not compensated (influence on retardation and axis orientation measurements will be discussed later) the retardation image (Fig. 1(b)) shows very interesting features. In the upper layers of the fovea no retardation (i.e., below the detection sensitivity of the instrument) can be observed (indicated by a rather constant color with depth; the small offset in the retardation values is introduced by birefringence of the anterior segment). On the other hand, retardation values of the last strong reflecting layer (part of the RPE/BM complex) change randomly in transversal direction (c.f., Fig. 1(e)). This indicates that light backscattered from within this layer is depolarized. (A direct measurement of the degree of polarization by a coherent detection technique as OCT is not possible) The polarization scrambling can be observed more clearly by the random axis orientation of this layer visible in Fig. 1(c). With our method it is now possible to differentiate between two layers of the RPE/BM complex differing either in tissue texture or in tissue structure. This might improve the association of layers observed in histological cross sections with layers of OCT images, especially to identify the two layers of the RPE/BM complex. Layers of the choroid (beneath the RPE/BM complex) appear, where visible, in rather well defined (i.e. not randomly varying) color, thus indicating that light transmitting the last layer of the RPE/BM complex is still in a defined polarization state (only light directly backscattered from the last layer of the RPE/BM complex is depolarized). In the lower left part of Fig. 1(b) an increase of retardation is visible which indicates an increased birefringence of the corresponding tissue (probably sclera) in this region. In Fig. 1(c) a predominant axis orientation in the whole image is visible which once again indicates an influence of anterior segment birefringence. (The slight change of the axis orientation as a function of lateral position might be introduced by a small deviation of the scanning beam pivot point from the center of the entrance pupil) Adjacent to the central part of the retina at the IPRL a slight increase in retardation can be observed which might be caused by Henle fibers.
Fig. 1. B-scan of human fovea in vivo (~5×1mm2) (a) intensity (the layers are labeled as follows: ILM, internal limiting membrane; NFL, nerve fiber layer; GCL, ganglion cell layer; IPL and OPL, inner and outer plexiform layer; INL and ONL, inner and outer nuclear layer; HF, Henle fiber layer; ELM, external limiting membrane; IPRL, interface between inner and outer segments of photoreceptor layer; RPE, retinal pigment epithelium), (b) retardation (c.f., color bar; δ=0° to 90°), (c) cumulative fast axis orientation (c.f., color bar; Θ=0° to 180°), (d) (e) (f) enlarged sections (x2) of (a), (b) and (c) (to avoid erroneous birefringence data values below a certain intensity threshold are displayed in grey in (b) (c) (e) (f))
Figure 2 shows a movie of different transversal (en face) sections corresponding to increasingly deeper positions within the tissue which was obtained from a 3D data set recorded from the fovea region of a volunteer in vivo. The 3D data set consists of 1000×100×100pixels and covers a volume of approximately 5×5×1.5mm3. Due to publication restraints each frame of the movie was converted down to 250×250 pixels and the total number of frames was reduced by a factor of 2. The upper left image of the movie shows a B-scan retrieved from the 3D-data set. A black line in the B-scan image indicates the position of the corresponding en-face images (upper right: intensity, lower left: cumulative axis orientation, lower right: retardation) in the movie. The black line in the en-face images indicates the B-scan position. Several layers of the fovea region can be identified in the B-scan image, although the lower depth resolution of the image (limited here by the number of recorded en-face images) prevents the visualization of the separation of the two layers of the RPE/BM complex. The first frames of the movie show the periphery of the fovea region corresponding to the section of the coherence plane with the retina. In the en-face intensity images the different layers of the retina appear as concentric rings and can be identified. Especially, the nerve fiber layer (innermost and strong reflecting ring) is clearly visible in the C-scan images corresponding to the anterior part of the retina. As expected the nerve fiber layer disappears in the central part of the fovea. As the coherence gate moves further through the tissue, strongly reflecting blood vessels become visible and cast shadows which appear as white lines in all the images below. The retardation and cumulative fast axis orientation images show constant values throughout all images corresponding to layers above the HF/OPL indicating no birefringence within these layers. As mentioned before, a constant offset in the retardation and fast axis orientation is introduced by birefringence of the anterior segment. As the coherence gate is moved further into the tissue, two strongly reflecting structures (visible as rings) corresponding to the IPRL and the RPE layer appear in these images. Different retardation values can be observed at each ring. The exterior ring of the RPE/BM complex shows randomly distributed retardation values as well as randomly distributed axis orientation values, indicating a depolarization of the light backscattered from this region. The interior ring (IPRL) shows low retardation at the periphery of the fovea.
Fig. 2. Frame no.34 of movie showing several en-face images of a 3D data set of a human fovea region in vivo at different depth positions. Upper left: intensity image of a B-scan (x-z) (black line corresponds to the depth position of the en face images), upper right: en face (x-y) intensity image, lower left: en face cumulative fast axis orientation image (c.f. color bar; Θ=0° to 180°), lower right: en face retardation image (c.f. color bar; δ=0° to 90°) The black lines correspond to the B-scan position (size 0.8MB) (The data set consists of a volume of 5×5×1.5mm3)
At the central part of the fovea an asymmetric retardation pattern (shaped like an hour glass) appears. Intensity images obtained from planes below the RPE/BM complex show structures of the choroid. The retardation images at these depths show low retardation (and a preservation of the polarization state) of light backscattered from within this region as already observed in Fig. 1(b).
The images shown in the movie of Fig. 2 represent sections of the coherence plane with the retina and observed patterns are difficult to interpret because the sectional planes are not parallel to the retinal layers. To improve the visualization of the retardation introduced by different layers we extracted the retardation values accumulated at different interfaces. To retrieve the surface of the retina we first reduced the amount of data points of the intensity images in x direction by a factor of 8 (corresponding to an averaging of adjacent A-lines in x direction). Then we searched the whole 3D-data set along each averaged A-line for the first value above a certain intensity threshold. This value was taken as the retinal surface. We interpolated between these surface positions in x direction to re-expand the data to the original data volume. Retardation and axis orientation corresponding to these surface voxels were taken as the retinal surface values (if the intensity of such a voxel was below a certain intensity threshold we displayed in the corresponding retardation and axis orientation images this voxel in grey). Figures 3(a) and 3(b) show retardation and cumulative fast axis orientation obtained at the surface of the fovea which corresponds to the retardation and fast axis orientation introduced by the anterior segment. Both images show rather constant values. The faint concentric rings present in each image are an artifact generated probably by a combination of a slight axial eye movement during measurement and an erroneous surface detection at the positions of the rings (the interpolation of surface points at positions where motion occurred can lead to sampling points off the real surface, where the signal can be very low; at these points, noise dominates which induces a color shift). A histogram of the distribution of retardation and cumulative fast axis orientation at the surface of the fovea is plotted in figs. 3(c) and 3(d). A mean retardation value of (18±11)° was calculated from the data (the error represents the standard deviation) which is in the range of values obtained from in vitro studies at the central part of the cornea [18, 36]. The calculation of a mean fast axis orientation deserves some care. As can be seen in Fig. 3(d) there is a constant offset in the distribution of the fast axis orientation (arises probably from noise introduced by a low signal in one of the polarization channels) which will result in a shift of the mean value. Therefore we have chosen an alternative way to determine the mean fast axis orientation by evaluating the peak of the distribution (the error is given by ½ of the FWHM of the distribution). We obtained a fast axis orientation of the anterior segment of (70±25)° which is equivalent to 70° nasally upward. For comparison we measured the retardation and axis orientation of the same eye with a commercially available SLP (Laser Diagnostic Technologies GDx VCC). With the commercial system we obtained a single pass retardation of 17.5° and a fast axis orientation of 55° of the anterior segment. The retardation value is in good agreement with our results. The mismatch in the fast axis orientation between the two methods might be caused by an offset in the measurement of the axis orientation with our system [27].
Fig. 3. Retardation (a) and cumulative fast axis orientation (b) at the surface of the fovea region. Each image covers an area of 5×5mm2. Histogram of retardation (c) and fast axis orientation (d) values obtained from (a) and (b), respectively.
Figures 4(a) and 4(b) show the retardation and cumulative fast axis orientation values obtained at the IPRL. The hour glass pattern in the central part of the fovea is clearly visible. A similar pattern was observed by SLP imaging and was attributed to an influence of corneal birefringence on the measurements [33]. It was assumed that the pattern arises by an overlap of two retarders: the anterior segment and the Henle fibers. While the anterior segment has a constant axis orientation, the Henle fibers are radially oriented and hence the corresponding axis orientation changes linearly with the azimuth angle. Based on Mueller calculus, the retardation δ for a two retarder combination (δ1, and δ2) can be expressed as [33]
with the axis orientation θ1 and θ2 of each retarder.
At positions with both axes parallel to each other the retardation corresponds to the sum of the two retarders (green color in Fig. 4(a)); at positions with both axes orthogonal to each other the retardation corresponds to the difference of the two retarders (dark blue in Fig. 4(a)). Therefore it is possible to obtain directly the fast axis orientation of the anterior segment from Fig. 4(a). It is oriented along the dark blue structure crossing the central area of Fig. 4(a) diagonally. We measured an axis orientation of 55° which corresponds to the value obtained with the SLP.
To verify the assumption of a two retarder combination we simulated the results of the experiment. In Jones formalism a general linear retarder can be represented by [27]
If we assume a vertically polarized incident beam on the interferometer [27], the Jones vector of the sample beam (two retarder combination) exiting the interferometer can be expressed as
with Mqwp the Jones matrix of a quarter wave plate (implemented in the experimental setup), MA and MH the Jones matrices of the anterior segment and the Henle fibers. R represents the sample reflectivity. Retardation and cumulative fast axis orientation can be calculated from Eqs. (2) and (3), respectively. Figures 4(c) and 4(d) show the simulated retardation and cumulative fast axis orientation of a two retarder combination with a retardation and fast axis orientation introduced by the anterior segment of δA=18° and θA=55° (c.f., Fig. (3) and the measurement with the SLP) and an assumed retardation introduced by the Henle fibers of θH=10° [37]. The fast axis orientation θH is assumed to increase linearly with the azimuth angle (from 0° to 360° in clockwise orientation) and to be perpendicular to the Henle fibers. The simulations are in good agreement with the experiment which justifies the assumption of a two retarder combination.
Fig. 4. Retardation (a) and cumulative fast axis orientation (b) at the IPRL of the fovea region (each image covers an area of 5×5mm2), and simulated retardation (c) and fast axis orientation (d) (S superior; I inferior; T temporal; N nasal)
Figure 5 shows a movie of consecutive transversal (en face) sections corresponding to increasingly deeper positions within the optic nerve head region of a volunteer recorded in vivo. The data set consists of a volume of approximately 5×5×2.5mm3. The arrangement of each data set in the movie is identical to the arrangement in Fig. 2. Several layers can be identified in the B-scan image. Notable is the increase of retardation as the coherence gate is moved deeper into the tissue. This is probably caused by the birefringence of the RNFL. To improve the visualization of these parameters we performed similar steps as for the surface detection. Here the averaged A-line was searched, starting from the maximum depth, along the -z direction for a value above a certain intensity threshold (which was set higher than for the surface detection). This value was assumed to correspond to the RPE (the strongest backscattering layer of the retina). From this point the last value above a certain intensity threshold along the -z direction was obtained which was attributed to the top layer of the RPE/BM complex. This procedure allowed us to extract the top layer of the RPE/BM complex near the optic nerve head (except for the lamina cribrosa).
Fig. 5. Frame no.21 of movie showing several en-face images of a 3D data set of a human retinal nerve head region in vivo at different depth positions. Upper left: (B-scan) intensity image (x-z) (black line corresponds to the depth position of the en face images), upper right: en face (x-y) intensity image, lower left: en face cumulative fast axis orientation image (c.f. color bar; Θ=0° to 180°), lower right: en face retardation image (c.f. color bar; δ=0° to 90°). The black lines correspond to the B-scan position (size 0.7MB). (The data set comprises a volume of 5×5×2.5mm3)
Fig. 6. Retardation (a) and cumulative fast axis orientation (b) at the top layer of the RPE of the nerve head region (each image covers an area of ~5×5mm2), and simulated retardation (c) and fast axis orientation (d) (S superior; I inferior; T temporal; N nasal)
Here the retardation image represents the combined retardation of anterior segment and RNFL. The pattern is similar to that observed in Fig. 4(a). The main difference here is that the retardation introduced by the nerve fiber layer is not constant along a circle around the nerve head [21] which somewhat distorts the hour glass pattern. In contrast to Fig. 4(b) the cumulative axis orientation in Fig. 6(b) shows a more pronounced pattern. The cumulative axis orientation oscillates in the whole range from 0° to 180° corresponding to the orientation of the fibers in the nerve fiber layer known from anatomy text books. A simulation (c.f., Figs. 6(c) and 6(d)) with the assumption of a constant retardation introduced by the nerve fiber layer (δN=34°) which exceeds the retardation introduced by the anterior segment yields a pattern roughly similar to that of Fig. 6(b). On the other hand the correlation between Figs 6(a) and 6(c) is rather poor because the retardation introduced by the RNFL is not constant in that region. The region inferior and superior to the nerve head causes higher retardation than nasal and temporal [21].
4. Discussion
We have demonstrated that our transversal scanning PS-OCT technique represents a powerful method to investigate the human retina and its polarizing properties in vivo. The axial resolution of the intensity images obtained by our system is sufficient to resolve several retinal layers, including layers that were previously seen only with ultrahigh resolution OCT [29]. This result is somewhat surprising because we used a conventional SLD with a vacuum coherence length of ~12 µm. It seems that visualization of these very thin layers (e.g., the external limiting membrane) depends more on the sampling density than on the coherence length of the light source. Our transversal scanning instrument provides, in the high-resolution B-scan mode, a rather high sampling density of 1600(x)×500(z) sampling points. Especially the transverse sampling density is considerably higher than that of conventional A-scan based time domain systems. The image acquisition time of 0.5 s for a B-scan image is also better than that of conventional time domain retinal OCT (~1 s in case of the Zeiss Humphrey OCT 3 for 400 A-scans), however, is slower than that of high-speed Fourier domain devices [8–10] which, however, have not yet been used for PS-OCT retinal imaging. The recording of a lower resolution 3D data set presently takes 10 s. While this is sufficiently short for imaging healthy subjects, it might be too slow for patients. The use of a resonant scanner might further reduce the measurement time [7].
An interesting finding that has not been previously reported is that the two layers of the RPE/BM complex show different properties with respect to the polarization state of backscattered light. While backscattering at the first layer maintains the polarization state of the light, the second layer acts as a depolarizer for backscattered light. On the other hand, light transmitted through these layers maintains its polarization state. The reason for the depolarized backscattering at the second layer of the RPE/BM complex is not yet clear. We speculate that backscattering at large, non-spherical particles [34], possibly melanocytes, might be the reason. If so, this effect might be useful for studying and diagnosing diseases that affect the RPE/BM complex, such as age related macular degeneration. However, further studies are necessary to confirm our speculation.
Apart from the depolarizing properties of the posterior layer of the RPE/BM complex, we studied properties of birefringent layers. The combined birefringent effects of the anterior segment and a layer within the retina located in the area around the fovea give rise to an hour-glass shaped pattern in transversal retardation images of the fovea region. A similar pattern was previously observed by SLP [33] and attributed to Henle’s fiber layer. However, since SLP does not provide depth resolved images, a final proof of this interpretation remained open. By comparing retardation patterns derived from the ILM and the IPRL, we were able to demonstrate that the second retarder is located within the retina, between ILM and IPRL. This clearly narrows the range of possible layers that might correspond to this retarder, and, given the azimutal variation of the optic axis of this retarder, is a very strong hint that this layer is indeed Henle’s fiber layer. Furthermore, a comparison of measured and simulated images enabled a quantitative determination of the retardation of this layer.
Another birefringent layer of the retina is the RNFL. This tissue is of great interest for glaucoma diagnostics [35]. The birefringence of the RNFL was recently measured quantitatively at different areas around the optic nerve head [21]. In that study, a different PS-OCT technique based on Stokes vector measurement was used. The Stokes vector technique is a differential technique that probes the sample with two or more polarization states per sample location successively. While our technique needs only one measurement per sample location to obtain retardation and axis orientation, it has the drawback that the results obtained are influenced by the birefringence of the anterior segment through which the retina is measured (the Stokes vector method does not suffer from that problem since it is a differential technique). This effect presently prevents quantitative measurements of the RNFL birefringence by our instrument, only qualitative information is available. It should be noted that a retardation of 90° introduced by the anterior segment would result in a linear polarization state incident on the retina. If this state is parallel to the fast axis of the retina, measurement of retinal retardation would be prevented. However, this should be no problem because a recent study carried out in more than 100 eyes has shown that the retardation introduced by the anterior segment is much smaller than 90° in all of the eyes examined [36]. Nevertheless, e.g., the general distribution of the RNFL around the optic nerve head (cf. Fig. 6b), which resembles the distribution of nerve fiber orientations, is quite convincing and corresponds to what is expected from anatomy textbooks. By individually compensating for anterior segment birefringence with a variable retarder, similarly to the method demonstrated very successfully in SLP [33], or with an appropriate software algorithm, taking advantage of the retardation and axis orientation measured at the retinal surface, this problem should be overcome and quantitative information on RNFL birefringence could become available with our technique.
5. Conclusion
We presented, to the best of our knowledge, the first three dimensional OCT images of retardation and (cumulative) fast axis orientation of two different regions (fovea, nerve head) of human retina in vivo. The longitudinal resolution in the B-scan mode of the system is sufficient to identify all layers which can be seen in ultra high resolution images. Furthermore, polarization sensitive images reveal a difference in the structure or texture of the two layers of the RPE/BM complex. The first of the two layers acts as a polarization preserving backscatterer while the second layer depolarizes backscattered light. This difference might be used for identifying the anatomical structures corresponding to these layers. The birefringence of the anterior segment causes a specific retardation pattern in the central part of the fovea which can be used as an indicator for anterior segment compensation. The change of orientation of the nerve fibers in the RNFL in the region around the nerve head could be visualized.
Acknowledgments
The authors wish to thank Dr. Q. Zhou, Laser Diagnostic Technologies, for performing GDx VCC measurements and L. Schachinger for technical assistance. Financial assistance from the Austrian Fonds zur Förderung der wissenschaftlichen Forschung (FWF-grant P16776-NO2) is acknowledged.
References and Links
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. G. Fujimoto, “Optical coherence tomography,” Science 254, 1178–1181 (1991). [CrossRef] [PubMed]
2. B. E. Bouma and G. J. Tearney, Handbook of Optical Coherence Tomography, (Marcel Dekker, New York2002).
3. A. F. Fercher and C. K. Hitzenberger, “Optical coherence tomography”, Chapter 4 in Progress in Optics44, Elsevier Science B.V. (2002)
4. C. A. Puliafito, M. R. Hee, C. P. Lin, E. Reichel, J. S. Schuman, J. S. Duker, J. A. Izatt, E. A. Swanson, and J. G. Fujimoto, “Imaging of macular diseases with optical coherence tomography,” Ophthalmology 102, 217–229 (1995) [PubMed]
5. P. Massin, C. Allouch, B. Haouchine, F. Metge, M. Paques, L. Tangui, A. Erginay, and A. Gaudric, “Optical coherence tomography of idiopathic maclular epiretinal membranes before and after surgery,” Am. J. Ophthalmology 130, 732–739 (2000) [CrossRef]
6. W. Drexler, H. Sattmann, B. Hermann, T. H. Ko, M. Stur, A. Unterhuber, C. Scholda, O. Findl, M. Wirtitsch, J. G. Fujimoto, and A. F. Fercher, “Enhanced visualization of macular pathology with use of ultrahigh-resolution optical coherence tomography,” Arch-Ophthalmology 121, 695–706 (2003) [CrossRef]
7. C. K. Hitzenberger, P. Trost, P. W. Lo, and Q. Zhou. “Three dimensional imaging of the human retina by high speed optical coherence tomography,” Opt. Express 11, 2753–2761 (2003) http://www.opticsexpress.org/abstract.cfm?URI=OPEX-11-21-2753 [CrossRef] [PubMed]
8. M. Wojtkowski, T. Bajraszewski, P. Targowski, and A. Kowalczyk, “Real-time in vivo imaging by high-speed spectral optical coherence tomography,” Opt. Lett. 28, 1745–1747 (2003) [CrossRef] [PubMed]
9. N. Nassif, B. Cense, B. H. Park, S. H. Yun, T. C. Chen, B. E. Bouma, G. J. Tearney, and J. F. de Boer, “In vivo human retinal imaging by ultrahigh-speed spectral domain optical coherence tomography,” Opt. Lett. 29, 480–482 (2004) [CrossRef] [PubMed]
10. R. A. Leitgeb, W. Drexler, A. Unterhuber, B. Hermann, T. Bajraszewski, T. Le, A. Stingl, and A. F. Fercher, “Ultrahigh resolution Fourier domain OCT,” Opt. Express 12, 2156–2165 (2004) http://www.opticsinfobase.org/abstract.cfm?id=79930 [CrossRef] [PubMed]
11. H. B. klein Brink and G. J. van Blockland, “Birefringence of the human foveal area assessed in vivo with Mueller-matrix ellipsometry,” J. Opt. Soc. Am. A 5, 49–57 (1988) [CrossRef]
12. A. W. Dreher, K. Reiter, and R. N. Weinreb, “Spatially resolved birefringence of the retinal nerve fiber layer assessed with a retinal laser ellipsometer,” Appl. Opti. 31, 3730–3735 (1992) [CrossRef]
13. 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. Am. B 9, 903–908 (1992) [CrossRef]
14. J. F. De Boer, T. E. Milner, M. J. C. van Gemert, and J. S. Nelson, “Two-dimensional birefringence imaging in biological tissue by polarization sensitive optical coherence tomography,” Opt. Lett. 22, 934–936 (1997) [CrossRef] [PubMed]
15. M. J. Everett, K. Schoenenberger, B. W. Colston Jr., and L. B. Da Silva, “Birefringence characterization of biological tissue by use of optical coherence tomography,” Opt. Lett. 23, 228–230 (1998) [CrossRef]
16. B. H. Park, M. C. Pierce, B. Cense, and J. F. de Boer, “Real-time multi-functional optical coherence tomography, ” Opt. Express 11, 782–793 (2003) http://www.opticsexpress.org/abstract.cfm?URI=OPEX-11-7-782 [CrossRef] [PubMed]
17. M. Pircher, E. Götzinger, R. Leitgeb, and C. K. Hitzenberger: “Three dimensional polarization sensitive OCT of human skin in vivo,” Opt. Express 12, 3236–3244 (2004) http://www.opticsexpress.org/abstract.cfm?URI=OPEX-12-14-3236 [CrossRef] [PubMed]
18. E. Goetzinger, M. Pircher, M. Sticker, A. F. Fercher, and C. K. Hitzenberger, “Measurement and imaging of birefringent properties of the human cornea with phase-resolved polarization-sensitive optical coherence tomography,” J. Biomed. Opt. 9, 94–102 (2004) [CrossRef]
19. B. Cense, T. C. Chen, B. H. Park, M. C. Pierce, and J. F. de Boer, “In vivo depth-resolved birefringence measurements of the human retinal nerve fiber layer by polarization sensitive optical coherence tomography,” Opt. Lett. 27, 1610–1612 (2002) [CrossRef]
20. M.G. Ducros, J. D. Marsack, H.G. Rylander III, S. L. Thomsen, and T. E. Milner“Primate retina imaging with polarization sensitive optical coherence tomography,” J. Opt. Soc. Am. A 18, 2945–2956 (2001) [CrossRef]
21. B. Cense, T. C. Chen, B. H. Park, M. C. Pierce, and J. F. de Boer, “Thickness and birefringence of healthy retinal nerve fiber layer tissue measured with polarization sensitive optical coherence tomography,” Invest. Ophthalmol. Vis. Sci. 45, 2606–2612 (2004) [CrossRef] [PubMed]
22. A. G. Podoleanu, G. M. Dobre, and D. A. Jackson, “En-face coherence imaging using galvanometer scanner modulation, ” Opt. Lett. 23, 147–149 (1998) [CrossRef]
23. A. G. Podoleanu, M. Seeger, G. M. Dobre, D. J. Webb, D. A. Jackson, and F. Fitzke, “Transversal and longitudinal images from the retina of the living eye using low coherence reflectometry, ” J. Biomed. Opt. 3, 12–20 (1998) [CrossRef] [PubMed]
24. R. G. Cucu, A. G. Podoleanu, R. B. Rosen, A. C. Boxer, and D. A. Jackson, “En face polarization sensitive optical coherence tomography,” Proc. SPIE 5140, 113–119/2003) [CrossRef]
25. A. G. Podoleanu, R. G. Cucu, and R. B. Rosen, “In vivo T-scan based polarization sensitive OCT of the optic nerve,” Proc. SPIE 5316, 300–305 (2004) [CrossRef]
26. M. Pircher, E. Goetzinger, R. Leitgeb, and C. K. Hitzenberger, “Transversal phase resolved polarization sensitive optical coherence tomography,” J. Phys. Med. Biol. 49, 1257–1263 (2004) [CrossRef]
27. C. K. Hitzenberger, E. Goetzinger, M. Sticker, M. Pircher, and A. F. Fercher, “Measurement and imaging of birefringence and optic axis orientation by phase resolved polarization sensitive optical coherence tomography,” Opt. Express 9, 780–790 (2001) http://www.opticsexpress.org/abstract.cfm?URI=OPEX-9-13-780 [CrossRef] [PubMed]
28. American National Standards Institute: “American National Standard for Safe Use of Lasers,” ANSI Z 136.1-2000. Orlando, Laser Institute of America, 45–49 (2000)
29. W. Drexler, U. Morgner, R. K. Ghanta, F. X. Kaertner, J. S. Schuhman, and J. G. Fujimoto, “Ultrahigh-resolution ophthalmic optical coherence tomography,” Nature Med. 7, 502–507 (2001) [CrossRef] [PubMed]
30. A. Baumgartner, C. K. Hitzenberger, H. Sattmann, W. Drexler, and A. F. Fercher, “Signal and resolution enhancements in dual beam optical coherence tomography of the human eye,” J. Biomed. Opt. 3, 45–54 (1998) [CrossRef] [PubMed]
31. B. Cense, N. A. Nassif, T. C. Chen, M. C. Pierce, S.-H. Yun, B. H. Park, B. E. Bouma, G. J. Tearney, and J. F. de Boer, “Ultrahigh-resolution high-speed retinal imaging using spectral-domain optical coherence tomography,” Opt. Express 12. 2435–2447 (2004) http://www.opticsexpress.org/abstract.cfm?URI=OPEX-12-11-2435 [CrossRef] [PubMed]
32. M. Wojtkowski, V. J. Srinivasan, T. H. Ko, J. G. Fujimoto, A. Kowalcyk, and J. S. Duker, “Ultrahigh-resolution, high-speed, Fourier domain optical coherence tomography and methods for dispersion compensation,” Opt. Express 12, 2404–2422 (2004) http://www.opticsexpress.org/abstract.cfm?URI=OPEX-12-11-2404 [CrossRef] [PubMed]
33. Q. Zhou and R. N. Weinreb, “Individualized compensation of anterior segment birefringence during laser polarimetry,” Invest. Ophthalmol. Vis. Sci. 43, 2221–2228 (2002) [PubMed]
34. J. M. Schmitt and S. H. Xiang, “Cross-polarized backscatter in optical coherence tomography of biological tissue,” Opt. Lett. 23, 1060–2062 (1998) [CrossRef]
35. M. J. Greaney, D. C. Hoffman, D. F. Garway-Heath, M. Nakla, A. L. Coleman, and J. Caprioli, “Comparison of optic nerve imaging methods to distinguish normal eyes from those with Glaucoma,” Invest. Ophthalmol. Vis. Sci. 43, 140–145 (2002) [PubMed]
36. R. W. Knighton and X. R. Huang, “Linear birefringence of the central human cornea,” Invest. Ophthalmol. Vis. Sci. 43, 82–86 (2002) [PubMed]
37. H. B. klein Brink and G. J. van Blockland, “Birefringence of the human foveal area assessed in vivo with Mueller matrix ellipsometry,” J. Opt. Soc. Am. A. 5, 49–57 (1988) [CrossRef]
