Orthogonal dispersive spectral-domain optical coherence tomography Wen Bao, Zhihua Ding,* Peng Li, Zhiyan Chen, Yi Shen, and Chuan Wang 1
State Key Lab of Modern Optical Instrumentation, Zhejiang University, 38 Zheda Rd., Hangzhou 310027, China *[email protected]
Abstract: Ultrahigh depth range spectral domain optical coherence tomography (SDOCT) can be realized based on the orthogonal dispersive spectrometer consisted by a high spectral resolution virtually-imaged phased array (VIPA) and a low spectral resolution grating. However, two critical issues result in the challenge of obtaining desirable one-dimensional (1-D) spectra from the recorded twodimensional (2-D) orthogonal spectra for high-quality OD-SDOCT imaging. One is the wavenumber mapping errors and the other is the periodic intensity modulations. The paper proposes a method for desirable reconstruction of 1-D spectra from the recorded 2-D orthogonal spectra. A sample etalon with identical parameters to the dispersive VIPA is used to determine the free spectrum range (FSR) of the VIPA, and spectral phases from two reflecting mirrors are further applied for broadband wavenumber calibration. The cascading of column spectra are performed from interval of four lines of column spectra, and four records of cascaded 1-D spectra are obtained and then averaged to alleviate the periodic intensity modulations. Broadband 1-D spectra are thus reconstructed with an ultrahigh spectral resolution. To demonstrate the feasibility of the proposed method, three typical samples are imaged by the OD-SDOCT system. ©2014 Optical Society of America OCIS codes: (170.4500) Optical coherence tomography; (150.1488) Calibration.
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 et, “Optical coherence tomography,” Science 254(5035), 1178–1181 (1991). 2. W. Drexler and J. G. Fujimoto, Optical Coherence Tomography: Technology and Applications (Springer, 2008). 3. J. Jungwirth, B. Baumann, M. Pircher, E. Götzinger, and C. K. Hitzenberger, “Extended in vivo anterior eye-segment imaging with full-range complex spectral domain optical coherence tomography,” J. Biomed. Opt. 14(5), 050501 (2009). 4. I. Grulkowski, M. Gora, M. Szkulmowski, I. Gorczynska, D. Szlag, S. Marcos, A. Kowalczyk, and M. Wojtkowski, “Anterior segment imaging with Spectral OCT system using a high-speed CMOS camera,” Opt. Express 17(6), 4842–4858 (2009). 5. P. Li, L. An, G. Lan, M. Johnstone, D. Malchow, and R. K. Wang, “Extended imaging depth to 12 mm for 1050-nm spectral domain optical coherence tomography for imaging the whole anterior segment of the human eye at 120-kHz A-scan rate,” J. Biomed. Opt. 18(1), 016012 (2013). 6. Z. Wang, Z. Yuan, H. Wang, and Y. Pan, “Increasing the imaging depth of spectral-domain OCT by using interpixel shift technique,” Opt. Express 14(16), 7014–7023 (2006). 7. T. Bajraszewski, M. Wojtkowski, M. Szkulmowski, A. Szkulmowska, R. Huber, and A. Kowalczyk, “Improved spectral optical coherence tomography using optical frequency comb,” Opt. Express 16(6), 4163–4176 (2008). 8. I. Grulkowski, J. J. Liu, B. Potsaid, V. Jayaraman, J. Jiang, J. G. Fujimoto, and A. E. Cable, “Highprecision, high-accuracy ultralong-range swept-source optical coherence tomography using vertical cavity surface emitting laser light source,” Opt. Lett. 38(5), 673–675 (2013). 9. I. Grulkowski, J. J. Liu, B. Potsaid, V. Jayaraman, C. D. Lu, J. Jiang, A. E. Cable, J. S. Duker, and J. G. Fujimoto, “Retinal, anterior segment and full eye imaging using ultrahigh speed swept source OCT with vertical-cavity surface emitting lasers,” Biomed. Opt. Express 3(11), 2733–2751 (2012). 10. C. Wang, Z. Ding, S. Mei, H. Yu, W. Hong, Y. Yan, and W. Shen, “Ultralong-range phase imaging with orthogonal dispersive spectral-domain optical coherence tomography,” Opt. Lett. 37(21), 4555–4557 (2012).
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10081
11. M. Shirasaki, “Large angular dispersion by a virtually imaged phased array and its application to a wavelength demultiplexer,” Opt. Lett. 21(5), 366–368 (1996). 12. S. A. Diddams, L. Hollberg, and V. Mbele, “Molecular fingerprinting with the resolved modes of a femtosecond laser frequency comb,” Nature 445(7128), 627–630 (2007). 13. S. A. van den Berg, S. T. Persijn, G. J. Kok, M. G. Zeitouny, and N. Bhattacharya, “Many-wavelength interferometry with thousands of lasers for absolute distance measurement,” Phys. Rev. Lett. 108(18), 183901 (2012). 14. S. Xiao and A. M. Weiner, “2-D wavelength demultiplexer with potential for >/= 1000 channels in the Cband,” Opt. Express 12(13), 2895–2902 (2004). 15. S. Wang, S. Xiao, and A. Weiner, “Broadband, high spectral resolution 2-D wavelength-parallel polarimeter for Dense WDM systems,” Opt. Express 13(23), 9374–9380 (2005). 16. A. Reyes-Reyes, M. Zeitouny, E. van Mastrigt, S. Persijn, N. Bhattacharya, and H. Urbach, “Cavityenhanced direct frequency comb spectroscopy,” in International Commission for Optics (2011), pp. 80112O1–80112O7. 17. L. Yang, “Analytical treatment of virtual image phase array.” in Proceedings of the Optical Fiber Communication Conference (2002), pp. 321–322. 18. S. Xiao, A. M. Weiner, and C. Lin, “Experimental and theoretical study of hyperfine WDM demultiplexer performance using the virtually imaged phased-array (VIPA),” J. Lightwave Technol. 23(3), 1456–1467 (2005). 19. A. Mokhtari and A. A. Shishegar, “Rigorous vectorial Gaussian beam modeling of spectral dispersing performance of virtually imaged phased arrays,” J. Opt. Soc. Am. B 26(2), 272–278 (2009). 20. S. Xiao, A. M. Weiner, and C. Lin, “A dispersion law for virtually imaged phased-array spectral dispersers based on paraxial wave theory,” IEEE J. Quantum Electron. 40(4), 420–426 (2004). 21. A. Weiner, “Femtosecond pulse shaping using spatial light modulators,” Rev. Sci. Instrum. 71(5), 1929– 1960 (2000). 22. J. Y. Lee and D. Y. Kim, “Versatile chromatic dispersion measurement of a single mode fiber using spectral white light interferometry,” Opt. Express 14(24), 11608–11615 (2006).
1. Introduction Fourier domain optical coherence tomography (FDOCT) is an imaging modality that provides cross-sectional images of tissue/material microstructures by spectral analysis of the low-coherence interference fringe pattern [1, 2]. One typical implementation of the FDOCT is spectral domain OCT (SDOCT), where a grating is used as a dispersive component working together with one-dimensional (1-D) line-scan camera. Such a configuration of the SDOCT achieves moderate spectral resolution and sampling rate, limiting the imaging range to several millimeters [2]. To extend the imaging range, efforts have been dedicated to improving the spatial frequency of the dispersive grating, reducing the pixel size and increasing the pixel number of the line-scan camera [3–5]. Alternative approaches such as inter pixel shifting [6] and optical frequency comb [7] have also been proposed to extend the imaging range in SDOCT. Inter pixel shifting is a method to increase the sampling rate and hence extend the imaging range, but unstable and slow due to mechanical movement of the camera [6]. By adopting an optical frequency comb in SDOCT, the spectrum sampling function is determined by the optical frequency comb rather than the CCD pixel size. Therefore, spectral resolution in SDOCT is greatly enhanced and an improved performance is achieved [7]. Another typical implementation of the FDOCT is swept source OCT (SSOCT), where the spectral resolution is mainly determined by the instantaneous bandwidth of the light from the swept source. A SSOCT based on a vertical-cavity surface-emitting laser (VCSEL) is recently demonstrated to enable imaging ranges from a few centimeters up to meters [8, 9]. In our previously work [10], an orthogonal dispersive SDOCT (OD-SDOCT) system is developed, which realizes an imaging depth over 80 mm, the longest depth range ever achieved by SDOCT. In the proposed OD-SDOCT system, the spectra were dispersed into 2-D space by a high spectral resolution virtually-imaged phased array (VIPA) [11] and a low spectral resolution diffraction grating, and recorded by a 2-D camera. The 2-D orthogonal dispersed spectra were then cascaded to form 1-D spectra with an ultrahigh spectral resolution of 2 pm [10]. However, to obtain desirable 1-D spectra for high-quality ODSDOCT imaging remains a great challenge. Such a challenge is also met in cases where VIPA-grating spectrometers are involved such as in molecular fingerprinting [12], distance measurement [13], wavelength demultiplexing [14, 15] and spectral analysis [16]. Two critical issues result in the challenge of obtaining desirable 1-D spectra for highquality OD-SDOCT imaging. The first issue to be resolved is the wavenumber mapping #207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10082
errors. Without a dedicated spectrum calibration procedure, the reconstructed OCT image is prone to be degraded with artifacts. An optical spectrum analyzer can be employed for spectrum calibration [14], but difficulty arises if the spectral resolution of the optical spectrum analyzer is limited. Laser frequency comb can be adopted for spectrum calibration, but at the cost of either in narrow bandwidth or in limited spectral resolution [12, 13]. Absorption line of an accurately known molecule such as carbon dioxide [16], or a known light source with an extremely narrow bandwidth [14], can also be used for spectrum calibration, but the requirement of broadband spectrum calibration make such approaches impractical. The second issue is the periodic intensity modulations. Intensity modulations over wavelength due to VIPA has been analytic treated by several groups. Yang [17] proposed a 2-D Gaussian beam as the periodic filter to model VIPA and calculated the maxima of passbands as a function of diffraction angle. Xiao et al [18] presented an analytical expression for the passband-response of VIPA demultiplexer. Mokhtari et al [19] formulated a 3-D vectorial Gaussian beam framework for VIPA. However, to our knowledge, the effect of intensity modulations on the cascaded 1-D spectra and the OCT image has never been touched. It is known that FFT operation is sensitive to periodic signal. Without an appropriate method to alleviate these periodic intensity modulations in the cascaded 1-D spectra, artifacts must appear in the OCT image. The aim of this paper is to achieve high-quality OCT imaging by the OD-SDOCT system. To this end, a method for reconstruction of 1-D spectra from the recorded 2-D orthogonal spectra is proposed. The method relies on a three-step procedure to reduce the wavenumber mapping errors and the periodic intensity modulations. To demonstrate the feasibility of the proposed method for desirable reconstruction of 1-D spectra and hence high-quality OCT imaging, three typical samples are imaged by OD-SDOCT system and OCT images are reconstructed by the proposed method. 2. Method
Fig. 1. (a) Schematic of the OD-SDOCT system based on a VIPA-grating spectrometer. (b) Detailed layout of the dispersive VIPA-grating spectrometer.
Figure 1 presents the schematic diagram of the proposed OD-SDOCT system based on a VIPA-grating spectrometer. In the system, a super luminescent diode (SLD 371-HP, Superlum Diodes Ltd) with a FWHM bandwidth of 45 nm centered at λ0 = 835 nm is used #207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10083
as the low-coherence source. The emitted light from SLD is coupled into a fiber-based Michelson interferometer via a broadband optical circulator. A 90/10 fiber coupler is used as the beam splitter. In the sample arm, the light is directed by a pair of X-Y galvanometer mirrors (6215H, Cambridge Technology) and then focused onto the sample by an objective lens with a focal length of 100 mm. The interference light returning from the reference arm and the sample arm is detected by a VIPA-grating based spectrometer, which is designed to disperse a broadband of 30 nm with an ultrahigh spectral resolution of 2 pm. In the spectrometer, the interference light is first collimated by a collimator (beam-waist radius W = 6.5 μm) and focused into a line by a cylindrical lens (Thorlabs, LJ1653L1-B, fc = 200 mm). The line-focused beam is incident onto the entrance window of a VIPA at an angle (θi) of 3°. The VIPA is a custom-built plane-parallel solid etalon, of which the back surface is coated with a partially reflective film (r ≈95%) and the front surface is coated with a reflective film (R ≈100%) except for an uncoated area used as the entrance window. The FSR of the VIPA is ~0.1 nm, mainly determined by its refractive index (n = 1.50) and geometrical thickness (t = 2.40 mm). The light after the VIPA is incident onto a grating (Wasatch Photonics, d = 1/1200 mm) with an angle (θg) of 30.1°. The orientation of the grating groove is normal to the focused line for orthogonal dispersion. The output spectra after the grating are focused by an imaging lens (Thorlabs, AC508-100-B, f = 200 mm) onto a 2-D CCD (UNIQ UP1830CL-12B, 1024 × 1024 pixels, pixel size of 6.45 μm, frame rate of 30Hz), as shown in Fig. 2. The spectral data fetched by the CCD are transferred to a computer via a high-speed frame grabber board for further data processing.
Fig. 2. Orthogonal spectra distribution on the CCD plane. The colorful lines indicate the dispersed spectra and the gridding represents pixels of the CCD.
The intensity distribution on the CCD plane under coordinate system x-y for one particular wavenumber is given by [12, 15, 20, 21]
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10084
fc 2 y 2
f 2W 2 2 2 kΔ (1 − Rr ) + 4 Rr sin ( ) 2
I CCD ( x, y , k ) ∝ I interference ( k ) ⋅ exp −2
1
⋅
x − α c(k − k0 ) 2 , 2 f / kW
⋅ exp −
(1) where, k is the wavenumber, I interference ( k ) is 1-D spectra of the interference light, Δ = 2nt cos(θin ) −
2t tan(θin ) cos(θi ) y t cos(θin ) y 2 − is the VIPA dispersion parameter, f nf 2
θin = θ i / n is the refraction angle inside the VIPA, α = 2π f / cdk 2 cos(θ g ) is the grating parameter, and k0 is the center wavenumber. If we simplify the spectrum position as the position where the peaked intensity is achieved using Eq. (1), then the spectrum position is determined by x peak ( k ) = α c( k − k0 ), y peak ( k ) = −
ntf tan(θ in ) cos(θ i ) ±
( ntf
(2)
tan(θ in ) cos(θ i ) ) + ntf cos(θ in ) ⋅ ( 2 nt cos(θ in ) − 2 mπ / k ) 2
2
t cos(θ in )
.
(3) Based on Eqs. (2) and (3), the recorded intensities at pixels corresponding to positions (x peak, y peak) are taken as the 2-D spectra to be used to cascade the 1-D interference spectra I ( k ) [10]. However, due to intensity modulations over wavelength caused by VIPA, the spectra recorded by the CCD are not ideal samplings of the spectra of the interference light. From Eqs. (1) to (3), the cascaded 1-D spectra can be expressed by f 2 y 2 (k ) I ( k ) = I int erference ( k ) ⋅ exp −2 c peak . f 2W 2
(4)
f 2 y 2 (k ) Here, exp −2 c peak is the intensity modulations function caused by VIPA. f 2W 2 Evidently, this intensity modulations functions periodically over all columns of the recorded 2-D spectra. Without an appropriate method to alleviate these periodic intensity modulations of the cascaded 1-D spectra, artifacts must appear in the OCT image. What is more, due to uncertainty in determination of the FSR-cutoffs, wavenumber offsets are unavoidably introduced during spectra cascading. Besides, wavenumbers are usually unevenly spaced over the 2-D CCD camera. Both wavenumber offsets and unevenly distribution could introduce noticeable wavenumber mapping errors in the cascaded 1-D spectra. Without dedicated spectrum calibration, the OCT image is prone to be degraded with artifacts. Therefore, reducing wavenumber mapping errors and periodic intensity modulations are essential to obtain high-quality OCT imaging by the OD-SDOCT system. In this paper, the method based on a three-step procedure including 1) FSR-cutoffs determination by a sample etalon, 2) wavenumber calibration by interference spectral phase from two reflecting mirrors, and 3) alleviation of periodic intensity modulations through averaging is proposed for reconstruction of 1-D spectra from the recorded 2-D orthogonal spectra. In the first step, a glass plate with the same parameters (thickness, refractive index and illumination geometry) as the dispersive solid etalon (VIPA) is used as the sample etalon, and the reference arm is further blocked to form a common-path configuration of the ODSDOCT system. The matching of the sample etalon and the dispersive etalon results in interference spectra with intensity modulations along column direction at a period equal to the FSR of the VIPA. This means the wavenumber intervals between adjacent positions of intensity maxima along column direction are exactly matched to the FSR of the VIPA.
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10085
Hence, FSR-cutoffs are directly appeared in the 2-D CCD and can be determined immediately. In the second step, the blocking in the reference arm is removed and a reflecting mirror is taken as the sample. Interference fringes generated by above two reflecting mirrors with an appropriate optical path difference (OPD) are then recorded by the 2-D CCD camera. Due to a linear relationship between spectral phase and wavenumber, spectral phases can be used for broadband wavenumber calibration. Phase retrieval is based on the basic principle that the phase difference between adjacent fringe peaks or valleys is 2π, and quadratic polynomial fitting and interpolation are further performed [22]. It is known to us that higher frequency of the pattern will make the fitting and interpolation more accurate, while too high frequency of the pattern will degrade the contrast of the pattern. Therefore the OPD should be adjusted to an appropriate value to obtain a suitable pattern for accurate wavenumber calibration. After FSR-cutoffs determination and wavenumber calibration, the third step of alleviation of periodic intensity modulations is done. The cascading of column spectra are performed from interval of four lines of column spectra, and four records of 1-D spectra are obtained from the 2-D spectra. As the wavenumbers between neighboring lines of column spectra are shifted along the dispersion direction of the VIPA, intensity modulations caused by VIPA can hence be alleviated by averaging of these four records of 1-D spectra. 3. Results
To determine FSR-cutoffs of the VIPA, a sample etalon with parameters identical to the dispersive etalon is measured by the OD-SDOCT system with the reference arm blocked. Figure 3 shows the resulted interference pattern on the 2-D CCD camera. Based on the intensity values of the recorded 2-D spectra, five solid white lines along the peaks and four dashed white lines along the valleys are linearly fitted and overlaid on these fringes. Representatively, repeated wavenumber positions corresponding to different spectral orders of the VIPA are highlighted by five yellow circles. Due to a rotation of the recoded system x’-y’ relative to the orthogonal dispersive system x-y, the VIPA dispersive direction corresponds to the solid blue line going through these circle-centers. Evidently, four complete FSRs bounded by the five solid lines are realized in the system.
Fig. 3. Interference pattern resulted from sample etalon with parameters identical to the dispersive etalon under common-path configuration of the OD-SDOCT system.
To conduct wavenumber calibration based on spectral phase, interference spectra between two reflective mirrors with an OPD of 30.13 mm is measured by the ODSDOCT system. The 30.13mm is calculated based on λ0 2 ⋅ Ppeak 4 N δλ , where Ppeak is the peak position in sample points, N is the total sample number. The recorded interference pattern is shown in Fig. 4(a). After subtracting the DC background shown in Fig. 4(b), the resulted interference fringes shown in Fig. 4(c) is used for phase retrieval
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10086
based on the principle of 2π phase difference between adjacent fringe peaks or valleys. One typical column spectra fetched at the column pointed by the arrow shown in Fig. 4(c) is given in Fig. 5(a), where two red lines indicate the column range limited by up and down boundaries of the four FSR-cutoffs. There are 35 peaks and 35 valleys appeared in this range, corresponding to 70 pixels with assigned phase stepped from 0 to 69π. These 70 pixels points with assigned phases are then used to calculate the interpolated phases for all the other pixels within this range. Consequently, the corresponding nonlinear phase curve is changed to be a linear phase curve as shown in Fig. 5(c). Based on the obtained phases for wavenumber calibration, the non-uniformly spaced spectra shown in Fig. 5(a) are hence interpolated to be the uniform spaced spectra shown in Fig. 5(b).
Fig. 4. Interference pattern resulting from two reflective mirrors with an OPD of 30.13 mm. (a) original pattern, (b) DC background, (c) the pattern after removing the DC background.
Fig. 5. One typical column of spectra fetched from Fig. 4(c) before (a) and after wavenumber calibration (b), and their corresponding spectral phase curves(c).
To reduce the periodic intensity modulations, averaging of four cascaded 1-D spectra are conducted after phase calibration. As an example, the feasibility of this alleviation is illustrated in Fig. 6 by the obtained spectra from above two reflecting mirrors. After cascading of column spectra from interval of four lines of column spectra in the 2-D spectra, four records of cascaded 1-D spectra are obtained as shown in Fig. 6(a). Evidently, periodic intensity modulations with the same frequency are available in these four records. Fortunately, there exists a fixed delay corresponding to the FSR of the VIPA between these records, and this can be used to alleviate the periodic intensity modulations caused by VIPA. A smoothed spectra without noticeable intensity modulations as shown in Fig. 6(b) is obtained by averaging of these four records. The FFT results of the 1-D cascaded spectra before and after averaging are shown in Figs. 6(c) and 6(d), respectively. The reflecting signal is ideally reconstructed in Fig. 6(d) while it is hardly discernable in Fig. 6(c), and the peak signal to noise ratio is measured to be 115dB.
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10087
Fig. 6. Four records of the cascaded 1-D spectra (a) and the averaged spectra (b), part of the FFT results from one record of the cascaded spectra (c) and the averaged spectra (d).
To validate the feasibility of the method based on the three-step procedure for obtaining desirable 1-D spectra and hence high-quality OD-SDOCT imaging, a dented metal plate and a model eye are taken as the samples. Figure 7(a) shows reconstructed OCT image of a dented metal plate without the three-step procedure applied. Evidently, this image is badly reconstructed with too many artifacts to be useable. In contrast, the reconstructed image shown in Fig. 7(b) is obtained with the three-step procedure applied, and the reflecting profile of the metal plate is correctly reconstructed without noticeable artifacts. A model eye (OEMI-7, Ocular Instruments, Inc., Bellevue, WA) with an overall length of 26 mm is also imaged by the OD-SDOCT system and the OCT image as shown in Fig. 8 is ideally reconstructed by the proposed method. Because of its curved shapes and low reflectivities, the back scattered light intensity from the model eye is weak and hence low SNR in OCT imaging. To demonstrate the ultralong-range of the OD-SDOCT system, a 102mm long steel post connected to a post holder base is further imaged and the reconstructed OCT image is shown in Fig. 9.
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10088
Fig. 7. OCT images of a dented metal plate with artifacts (a) and without artifacts (b).
Fig. 8. Imaging of a model eye by the OD-SDOCT system with the proposed method.
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10089
Fig. 9. Imaging of a 102mm long steel post connected to a post holder base by the ODSDOCT system.
4. Conclusion
We develop a method based on a three-step procedure to resolve the challenge of obtaining desirable 1-D spectra from the 2-D orthogonal spectra. The FSR of the VIPA is determined from the spectra of a sample etalon, and the broadband spectra are calibrated by the spectral phase. The cascading of column spectra are performed from interval of four lines of column spectra, and four records of 1-D spectra are obtained and further averaged to form a smoothed 1-D spectra. This smoothed 1-D spectra with ultrahigh spectra resolution are then used for OCT reconstruction. Three typical samples are imaged by the OD-SDOCT system and high-quality OD-SDOCT images are obtained by the proposed method. The developed OD-SDOCT inherits the merits of SDOCT in phase stability, and might be more suitable than the VCSEL-based SSOCT for functional extensions such as Doppler imaging. However, the VCSEL-based SSOCT has the advantage in imaging speed. Anyway, with faster frame rate of the 2-D detector, the ODSDOCT with flexible parameter settings is a promising ultralong-range imaging modality. Acknowledgments
The authors would like to acknowledge the financial supports from the Chinese Natural Science Foundation (61335003, 61275196, 61327007), Zhejiang Province Science and Technology Grant (2012C33031) and the Fundamental Research Funds for the Central Universities (2014QNA5017).
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10090
State Key Lab of Modern Optical Instrumentation, Zhejiang University, 38 Zheda Rd., Hangzhou 310027, China *[email protected]
Abstract: Ultrahigh depth range spectral domain optical coherence tomography (SDOCT) can be realized based on the orthogonal dispersive spectrometer consisted by a high spectral resolution virtually-imaged phased array (VIPA) and a low spectral resolution grating. However, two critical issues result in the challenge of obtaining desirable one-dimensional (1-D) spectra from the recorded twodimensional (2-D) orthogonal spectra for high-quality OD-SDOCT imaging. One is the wavenumber mapping errors and the other is the periodic intensity modulations. The paper proposes a method for desirable reconstruction of 1-D spectra from the recorded 2-D orthogonal spectra. A sample etalon with identical parameters to the dispersive VIPA is used to determine the free spectrum range (FSR) of the VIPA, and spectral phases from two reflecting mirrors are further applied for broadband wavenumber calibration. The cascading of column spectra are performed from interval of four lines of column spectra, and four records of cascaded 1-D spectra are obtained and then averaged to alleviate the periodic intensity modulations. Broadband 1-D spectra are thus reconstructed with an ultrahigh spectral resolution. To demonstrate the feasibility of the proposed method, three typical samples are imaged by the OD-SDOCT system. ©2014 Optical Society of America OCIS codes: (170.4500) Optical coherence tomography; (150.1488) Calibration.
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 et, “Optical coherence tomography,” Science 254(5035), 1178–1181 (1991). 2. W. Drexler and J. G. Fujimoto, Optical Coherence Tomography: Technology and Applications (Springer, 2008). 3. J. Jungwirth, B. Baumann, M. Pircher, E. Götzinger, and C. K. Hitzenberger, “Extended in vivo anterior eye-segment imaging with full-range complex spectral domain optical coherence tomography,” J. Biomed. Opt. 14(5), 050501 (2009). 4. I. Grulkowski, M. Gora, M. Szkulmowski, I. Gorczynska, D. Szlag, S. Marcos, A. Kowalczyk, and M. Wojtkowski, “Anterior segment imaging with Spectral OCT system using a high-speed CMOS camera,” Opt. Express 17(6), 4842–4858 (2009). 5. P. Li, L. An, G. Lan, M. Johnstone, D. Malchow, and R. K. Wang, “Extended imaging depth to 12 mm for 1050-nm spectral domain optical coherence tomography for imaging the whole anterior segment of the human eye at 120-kHz A-scan rate,” J. Biomed. Opt. 18(1), 016012 (2013). 6. Z. Wang, Z. Yuan, H. Wang, and Y. Pan, “Increasing the imaging depth of spectral-domain OCT by using interpixel shift technique,” Opt. Express 14(16), 7014–7023 (2006). 7. T. Bajraszewski, M. Wojtkowski, M. Szkulmowski, A. Szkulmowska, R. Huber, and A. Kowalczyk, “Improved spectral optical coherence tomography using optical frequency comb,” Opt. Express 16(6), 4163–4176 (2008). 8. I. Grulkowski, J. J. Liu, B. Potsaid, V. Jayaraman, J. Jiang, J. G. Fujimoto, and A. E. Cable, “Highprecision, high-accuracy ultralong-range swept-source optical coherence tomography using vertical cavity surface emitting laser light source,” Opt. Lett. 38(5), 673–675 (2013). 9. I. Grulkowski, J. J. Liu, B. Potsaid, V. Jayaraman, C. D. Lu, J. Jiang, A. E. Cable, J. S. Duker, and J. G. Fujimoto, “Retinal, anterior segment and full eye imaging using ultrahigh speed swept source OCT with vertical-cavity surface emitting lasers,” Biomed. Opt. Express 3(11), 2733–2751 (2012). 10. C. Wang, Z. Ding, S. Mei, H. Yu, W. Hong, Y. Yan, and W. Shen, “Ultralong-range phase imaging with orthogonal dispersive spectral-domain optical coherence tomography,” Opt. Lett. 37(21), 4555–4557 (2012).
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10081
11. M. Shirasaki, “Large angular dispersion by a virtually imaged phased array and its application to a wavelength demultiplexer,” Opt. Lett. 21(5), 366–368 (1996). 12. S. A. Diddams, L. Hollberg, and V. Mbele, “Molecular fingerprinting with the resolved modes of a femtosecond laser frequency comb,” Nature 445(7128), 627–630 (2007). 13. S. A. van den Berg, S. T. Persijn, G. J. Kok, M. G. Zeitouny, and N. Bhattacharya, “Many-wavelength interferometry with thousands of lasers for absolute distance measurement,” Phys. Rev. Lett. 108(18), 183901 (2012). 14. S. Xiao and A. M. Weiner, “2-D wavelength demultiplexer with potential for >/= 1000 channels in the Cband,” Opt. Express 12(13), 2895–2902 (2004). 15. S. Wang, S. Xiao, and A. Weiner, “Broadband, high spectral resolution 2-D wavelength-parallel polarimeter for Dense WDM systems,” Opt. Express 13(23), 9374–9380 (2005). 16. A. Reyes-Reyes, M. Zeitouny, E. van Mastrigt, S. Persijn, N. Bhattacharya, and H. Urbach, “Cavityenhanced direct frequency comb spectroscopy,” in International Commission for Optics (2011), pp. 80112O1–80112O7. 17. L. Yang, “Analytical treatment of virtual image phase array.” in Proceedings of the Optical Fiber Communication Conference (2002), pp. 321–322. 18. S. Xiao, A. M. Weiner, and C. Lin, “Experimental and theoretical study of hyperfine WDM demultiplexer performance using the virtually imaged phased-array (VIPA),” J. Lightwave Technol. 23(3), 1456–1467 (2005). 19. A. Mokhtari and A. A. Shishegar, “Rigorous vectorial Gaussian beam modeling of spectral dispersing performance of virtually imaged phased arrays,” J. Opt. Soc. Am. B 26(2), 272–278 (2009). 20. S. Xiao, A. M. Weiner, and C. Lin, “A dispersion law for virtually imaged phased-array spectral dispersers based on paraxial wave theory,” IEEE J. Quantum Electron. 40(4), 420–426 (2004). 21. A. Weiner, “Femtosecond pulse shaping using spatial light modulators,” Rev. Sci. Instrum. 71(5), 1929– 1960 (2000). 22. J. Y. Lee and D. Y. Kim, “Versatile chromatic dispersion measurement of a single mode fiber using spectral white light interferometry,” Opt. Express 14(24), 11608–11615 (2006).
1. Introduction Fourier domain optical coherence tomography (FDOCT) is an imaging modality that provides cross-sectional images of tissue/material microstructures by spectral analysis of the low-coherence interference fringe pattern [1, 2]. One typical implementation of the FDOCT is spectral domain OCT (SDOCT), where a grating is used as a dispersive component working together with one-dimensional (1-D) line-scan camera. Such a configuration of the SDOCT achieves moderate spectral resolution and sampling rate, limiting the imaging range to several millimeters [2]. To extend the imaging range, efforts have been dedicated to improving the spatial frequency of the dispersive grating, reducing the pixel size and increasing the pixel number of the line-scan camera [3–5]. Alternative approaches such as inter pixel shifting [6] and optical frequency comb [7] have also been proposed to extend the imaging range in SDOCT. Inter pixel shifting is a method to increase the sampling rate and hence extend the imaging range, but unstable and slow due to mechanical movement of the camera [6]. By adopting an optical frequency comb in SDOCT, the spectrum sampling function is determined by the optical frequency comb rather than the CCD pixel size. Therefore, spectral resolution in SDOCT is greatly enhanced and an improved performance is achieved [7]. Another typical implementation of the FDOCT is swept source OCT (SSOCT), where the spectral resolution is mainly determined by the instantaneous bandwidth of the light from the swept source. A SSOCT based on a vertical-cavity surface-emitting laser (VCSEL) is recently demonstrated to enable imaging ranges from a few centimeters up to meters [8, 9]. In our previously work [10], an orthogonal dispersive SDOCT (OD-SDOCT) system is developed, which realizes an imaging depth over 80 mm, the longest depth range ever achieved by SDOCT. In the proposed OD-SDOCT system, the spectra were dispersed into 2-D space by a high spectral resolution virtually-imaged phased array (VIPA) [11] and a low spectral resolution diffraction grating, and recorded by a 2-D camera. The 2-D orthogonal dispersed spectra were then cascaded to form 1-D spectra with an ultrahigh spectral resolution of 2 pm [10]. However, to obtain desirable 1-D spectra for high-quality ODSDOCT imaging remains a great challenge. Such a challenge is also met in cases where VIPA-grating spectrometers are involved such as in molecular fingerprinting [12], distance measurement [13], wavelength demultiplexing [14, 15] and spectral analysis [16]. Two critical issues result in the challenge of obtaining desirable 1-D spectra for highquality OD-SDOCT imaging. The first issue to be resolved is the wavenumber mapping #207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10082
errors. Without a dedicated spectrum calibration procedure, the reconstructed OCT image is prone to be degraded with artifacts. An optical spectrum analyzer can be employed for spectrum calibration [14], but difficulty arises if the spectral resolution of the optical spectrum analyzer is limited. Laser frequency comb can be adopted for spectrum calibration, but at the cost of either in narrow bandwidth or in limited spectral resolution [12, 13]. Absorption line of an accurately known molecule such as carbon dioxide [16], or a known light source with an extremely narrow bandwidth [14], can also be used for spectrum calibration, but the requirement of broadband spectrum calibration make such approaches impractical. The second issue is the periodic intensity modulations. Intensity modulations over wavelength due to VIPA has been analytic treated by several groups. Yang [17] proposed a 2-D Gaussian beam as the periodic filter to model VIPA and calculated the maxima of passbands as a function of diffraction angle. Xiao et al [18] presented an analytical expression for the passband-response of VIPA demultiplexer. Mokhtari et al [19] formulated a 3-D vectorial Gaussian beam framework for VIPA. However, to our knowledge, the effect of intensity modulations on the cascaded 1-D spectra and the OCT image has never been touched. It is known that FFT operation is sensitive to periodic signal. Without an appropriate method to alleviate these periodic intensity modulations in the cascaded 1-D spectra, artifacts must appear in the OCT image. The aim of this paper is to achieve high-quality OCT imaging by the OD-SDOCT system. To this end, a method for reconstruction of 1-D spectra from the recorded 2-D orthogonal spectra is proposed. The method relies on a three-step procedure to reduce the wavenumber mapping errors and the periodic intensity modulations. To demonstrate the feasibility of the proposed method for desirable reconstruction of 1-D spectra and hence high-quality OCT imaging, three typical samples are imaged by OD-SDOCT system and OCT images are reconstructed by the proposed method. 2. Method
Fig. 1. (a) Schematic of the OD-SDOCT system based on a VIPA-grating spectrometer. (b) Detailed layout of the dispersive VIPA-grating spectrometer.
Figure 1 presents the schematic diagram of the proposed OD-SDOCT system based on a VIPA-grating spectrometer. In the system, a super luminescent diode (SLD 371-HP, Superlum Diodes Ltd) with a FWHM bandwidth of 45 nm centered at λ0 = 835 nm is used #207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10083
as the low-coherence source. The emitted light from SLD is coupled into a fiber-based Michelson interferometer via a broadband optical circulator. A 90/10 fiber coupler is used as the beam splitter. In the sample arm, the light is directed by a pair of X-Y galvanometer mirrors (6215H, Cambridge Technology) and then focused onto the sample by an objective lens with a focal length of 100 mm. The interference light returning from the reference arm and the sample arm is detected by a VIPA-grating based spectrometer, which is designed to disperse a broadband of 30 nm with an ultrahigh spectral resolution of 2 pm. In the spectrometer, the interference light is first collimated by a collimator (beam-waist radius W = 6.5 μm) and focused into a line by a cylindrical lens (Thorlabs, LJ1653L1-B, fc = 200 mm). The line-focused beam is incident onto the entrance window of a VIPA at an angle (θi) of 3°. The VIPA is a custom-built plane-parallel solid etalon, of which the back surface is coated with a partially reflective film (r ≈95%) and the front surface is coated with a reflective film (R ≈100%) except for an uncoated area used as the entrance window. The FSR of the VIPA is ~0.1 nm, mainly determined by its refractive index (n = 1.50) and geometrical thickness (t = 2.40 mm). The light after the VIPA is incident onto a grating (Wasatch Photonics, d = 1/1200 mm) with an angle (θg) of 30.1°. The orientation of the grating groove is normal to the focused line for orthogonal dispersion. The output spectra after the grating are focused by an imaging lens (Thorlabs, AC508-100-B, f = 200 mm) onto a 2-D CCD (UNIQ UP1830CL-12B, 1024 × 1024 pixels, pixel size of 6.45 μm, frame rate of 30Hz), as shown in Fig. 2. The spectral data fetched by the CCD are transferred to a computer via a high-speed frame grabber board for further data processing.
Fig. 2. Orthogonal spectra distribution on the CCD plane. The colorful lines indicate the dispersed spectra and the gridding represents pixels of the CCD.
The intensity distribution on the CCD plane under coordinate system x-y for one particular wavenumber is given by [12, 15, 20, 21]
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10084
fc 2 y 2
f 2W 2 2 2 kΔ (1 − Rr ) + 4 Rr sin ( ) 2
I CCD ( x, y , k ) ∝ I interference ( k ) ⋅ exp −2
1
⋅
x − α c(k − k0 ) 2 , 2 f / kW
⋅ exp −
(1) where, k is the wavenumber, I interference ( k ) is 1-D spectra of the interference light, Δ = 2nt cos(θin ) −
2t tan(θin ) cos(θi ) y t cos(θin ) y 2 − is the VIPA dispersion parameter, f nf 2
θin = θ i / n is the refraction angle inside the VIPA, α = 2π f / cdk 2 cos(θ g ) is the grating parameter, and k0 is the center wavenumber. If we simplify the spectrum position as the position where the peaked intensity is achieved using Eq. (1), then the spectrum position is determined by x peak ( k ) = α c( k − k0 ), y peak ( k ) = −
ntf tan(θ in ) cos(θ i ) ±
( ntf
(2)
tan(θ in ) cos(θ i ) ) + ntf cos(θ in ) ⋅ ( 2 nt cos(θ in ) − 2 mπ / k ) 2
2
t cos(θ in )
.
(3) Based on Eqs. (2) and (3), the recorded intensities at pixels corresponding to positions (x peak, y peak) are taken as the 2-D spectra to be used to cascade the 1-D interference spectra I ( k ) [10]. However, due to intensity modulations over wavelength caused by VIPA, the spectra recorded by the CCD are not ideal samplings of the spectra of the interference light. From Eqs. (1) to (3), the cascaded 1-D spectra can be expressed by f 2 y 2 (k ) I ( k ) = I int erference ( k ) ⋅ exp −2 c peak . f 2W 2
(4)
f 2 y 2 (k ) Here, exp −2 c peak is the intensity modulations function caused by VIPA. f 2W 2 Evidently, this intensity modulations functions periodically over all columns of the recorded 2-D spectra. Without an appropriate method to alleviate these periodic intensity modulations of the cascaded 1-D spectra, artifacts must appear in the OCT image. What is more, due to uncertainty in determination of the FSR-cutoffs, wavenumber offsets are unavoidably introduced during spectra cascading. Besides, wavenumbers are usually unevenly spaced over the 2-D CCD camera. Both wavenumber offsets and unevenly distribution could introduce noticeable wavenumber mapping errors in the cascaded 1-D spectra. Without dedicated spectrum calibration, the OCT image is prone to be degraded with artifacts. Therefore, reducing wavenumber mapping errors and periodic intensity modulations are essential to obtain high-quality OCT imaging by the OD-SDOCT system. In this paper, the method based on a three-step procedure including 1) FSR-cutoffs determination by a sample etalon, 2) wavenumber calibration by interference spectral phase from two reflecting mirrors, and 3) alleviation of periodic intensity modulations through averaging is proposed for reconstruction of 1-D spectra from the recorded 2-D orthogonal spectra. In the first step, a glass plate with the same parameters (thickness, refractive index and illumination geometry) as the dispersive solid etalon (VIPA) is used as the sample etalon, and the reference arm is further blocked to form a common-path configuration of the ODSDOCT system. The matching of the sample etalon and the dispersive etalon results in interference spectra with intensity modulations along column direction at a period equal to the FSR of the VIPA. This means the wavenumber intervals between adjacent positions of intensity maxima along column direction are exactly matched to the FSR of the VIPA.
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10085
Hence, FSR-cutoffs are directly appeared in the 2-D CCD and can be determined immediately. In the second step, the blocking in the reference arm is removed and a reflecting mirror is taken as the sample. Interference fringes generated by above two reflecting mirrors with an appropriate optical path difference (OPD) are then recorded by the 2-D CCD camera. Due to a linear relationship between spectral phase and wavenumber, spectral phases can be used for broadband wavenumber calibration. Phase retrieval is based on the basic principle that the phase difference between adjacent fringe peaks or valleys is 2π, and quadratic polynomial fitting and interpolation are further performed [22]. It is known to us that higher frequency of the pattern will make the fitting and interpolation more accurate, while too high frequency of the pattern will degrade the contrast of the pattern. Therefore the OPD should be adjusted to an appropriate value to obtain a suitable pattern for accurate wavenumber calibration. After FSR-cutoffs determination and wavenumber calibration, the third step of alleviation of periodic intensity modulations is done. The cascading of column spectra are performed from interval of four lines of column spectra, and four records of 1-D spectra are obtained from the 2-D spectra. As the wavenumbers between neighboring lines of column spectra are shifted along the dispersion direction of the VIPA, intensity modulations caused by VIPA can hence be alleviated by averaging of these four records of 1-D spectra. 3. Results
To determine FSR-cutoffs of the VIPA, a sample etalon with parameters identical to the dispersive etalon is measured by the OD-SDOCT system with the reference arm blocked. Figure 3 shows the resulted interference pattern on the 2-D CCD camera. Based on the intensity values of the recorded 2-D spectra, five solid white lines along the peaks and four dashed white lines along the valleys are linearly fitted and overlaid on these fringes. Representatively, repeated wavenumber positions corresponding to different spectral orders of the VIPA are highlighted by five yellow circles. Due to a rotation of the recoded system x’-y’ relative to the orthogonal dispersive system x-y, the VIPA dispersive direction corresponds to the solid blue line going through these circle-centers. Evidently, four complete FSRs bounded by the five solid lines are realized in the system.
Fig. 3. Interference pattern resulted from sample etalon with parameters identical to the dispersive etalon under common-path configuration of the OD-SDOCT system.
To conduct wavenumber calibration based on spectral phase, interference spectra between two reflective mirrors with an OPD of 30.13 mm is measured by the ODSDOCT system. The 30.13mm is calculated based on λ0 2 ⋅ Ppeak 4 N δλ , where Ppeak is the peak position in sample points, N is the total sample number. The recorded interference pattern is shown in Fig. 4(a). After subtracting the DC background shown in Fig. 4(b), the resulted interference fringes shown in Fig. 4(c) is used for phase retrieval
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10086
based on the principle of 2π phase difference between adjacent fringe peaks or valleys. One typical column spectra fetched at the column pointed by the arrow shown in Fig. 4(c) is given in Fig. 5(a), where two red lines indicate the column range limited by up and down boundaries of the four FSR-cutoffs. There are 35 peaks and 35 valleys appeared in this range, corresponding to 70 pixels with assigned phase stepped from 0 to 69π. These 70 pixels points with assigned phases are then used to calculate the interpolated phases for all the other pixels within this range. Consequently, the corresponding nonlinear phase curve is changed to be a linear phase curve as shown in Fig. 5(c). Based on the obtained phases for wavenumber calibration, the non-uniformly spaced spectra shown in Fig. 5(a) are hence interpolated to be the uniform spaced spectra shown in Fig. 5(b).
Fig. 4. Interference pattern resulting from two reflective mirrors with an OPD of 30.13 mm. (a) original pattern, (b) DC background, (c) the pattern after removing the DC background.
Fig. 5. One typical column of spectra fetched from Fig. 4(c) before (a) and after wavenumber calibration (b), and their corresponding spectral phase curves(c).
To reduce the periodic intensity modulations, averaging of four cascaded 1-D spectra are conducted after phase calibration. As an example, the feasibility of this alleviation is illustrated in Fig. 6 by the obtained spectra from above two reflecting mirrors. After cascading of column spectra from interval of four lines of column spectra in the 2-D spectra, four records of cascaded 1-D spectra are obtained as shown in Fig. 6(a). Evidently, periodic intensity modulations with the same frequency are available in these four records. Fortunately, there exists a fixed delay corresponding to the FSR of the VIPA between these records, and this can be used to alleviate the periodic intensity modulations caused by VIPA. A smoothed spectra without noticeable intensity modulations as shown in Fig. 6(b) is obtained by averaging of these four records. The FFT results of the 1-D cascaded spectra before and after averaging are shown in Figs. 6(c) and 6(d), respectively. The reflecting signal is ideally reconstructed in Fig. 6(d) while it is hardly discernable in Fig. 6(c), and the peak signal to noise ratio is measured to be 115dB.
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10087
Fig. 6. Four records of the cascaded 1-D spectra (a) and the averaged spectra (b), part of the FFT results from one record of the cascaded spectra (c) and the averaged spectra (d).
To validate the feasibility of the method based on the three-step procedure for obtaining desirable 1-D spectra and hence high-quality OD-SDOCT imaging, a dented metal plate and a model eye are taken as the samples. Figure 7(a) shows reconstructed OCT image of a dented metal plate without the three-step procedure applied. Evidently, this image is badly reconstructed with too many artifacts to be useable. In contrast, the reconstructed image shown in Fig. 7(b) is obtained with the three-step procedure applied, and the reflecting profile of the metal plate is correctly reconstructed without noticeable artifacts. A model eye (OEMI-7, Ocular Instruments, Inc., Bellevue, WA) with an overall length of 26 mm is also imaged by the OD-SDOCT system and the OCT image as shown in Fig. 8 is ideally reconstructed by the proposed method. Because of its curved shapes and low reflectivities, the back scattered light intensity from the model eye is weak and hence low SNR in OCT imaging. To demonstrate the ultralong-range of the OD-SDOCT system, a 102mm long steel post connected to a post holder base is further imaged and the reconstructed OCT image is shown in Fig. 9.
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10088
Fig. 7. OCT images of a dented metal plate with artifacts (a) and without artifacts (b).
Fig. 8. Imaging of a model eye by the OD-SDOCT system with the proposed method.
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10089
Fig. 9. Imaging of a 102mm long steel post connected to a post holder base by the ODSDOCT system.
4. Conclusion
We develop a method based on a three-step procedure to resolve the challenge of obtaining desirable 1-D spectra from the 2-D orthogonal spectra. The FSR of the VIPA is determined from the spectra of a sample etalon, and the broadband spectra are calibrated by the spectral phase. The cascading of column spectra are performed from interval of four lines of column spectra, and four records of 1-D spectra are obtained and further averaged to form a smoothed 1-D spectra. This smoothed 1-D spectra with ultrahigh spectra resolution are then used for OCT reconstruction. Three typical samples are imaged by the OD-SDOCT system and high-quality OD-SDOCT images are obtained by the proposed method. The developed OD-SDOCT inherits the merits of SDOCT in phase stability, and might be more suitable than the VCSEL-based SSOCT for functional extensions such as Doppler imaging. However, the VCSEL-based SSOCT has the advantage in imaging speed. Anyway, with faster frame rate of the 2-D detector, the ODSDOCT with flexible parameter settings is a promising ultralong-range imaging modality. Acknowledgments
The authors would like to acknowledge the financial supports from the Chinese Natural Science Foundation (61335003, 61275196, 61327007), Zhejiang Province Science and Technology Grant (2012C33031) and the Fundamental Research Funds for the Central Universities (2014QNA5017).
#207186 - $15.00 USD (C) 2014 OSA
Received 27 Feb 2014; revised 8 Apr 2014; accepted 9 Apr 2014; published 18 Apr 2014 21 April 2014 | Vol. 22, No. 8 | DOI:10.1364/OE.22.010081 | OPTICS EXPRESS 10090
![[Optical coherence tomography].](/assets/img/hold.png)
