Double common-path interferometer for flexible optical probe of optical coherence tomography

Transcription

1 Double ommon-path interferometer for flexible optial probe of optial oherene tomography Jae Seok Park, 1 Zhongping Chen, 1,2 Myung Yung Jeong, 1 and Chang-Seok Kim 1,* 1 Department of Cogno Mehatronis Engineering, World Class University Program, Pusan National University, Busan , South Korea 2 Bekman Laser Institute, Department of Biomedial Engineering, University of California, Irvine, California 92612, USA *kim@pusan.a.kr Abstrat: A flexible urled optial ord is useful for a ommon-path optial oherene tomography (OCT) system beause a bending-insensitive arbitrary length an be hosen for the endosopi imaging probe. However, there has been a ritial problem that the partial refletor needs to be plaed in between the sample and the objetive lens. It limits the struture design of optial probe and leads to a low transverse resolution OCT imaging. Instead of a onventional single ommon-path interferometer, we propose a novel double ommon-path interferometer onfiguration in order to generate an interferene signal that is independent of the optial distane between the partial refletor and sample. Due to the limitless tuning of the objetive distane, an objetive lens with a high numerial aperture (NA) up to 0.85 an be suessfully used for phase-sensitive optial oherene tomography to ahieve a 3-dimensional profile image of a transverse resolution of 0.7 μm. The intensity and phase terms of the interferene signal an be obtained simultaneously from a Fourier-domain mode loked swept laser soure for fast data aquisition with a phase stability of 979 pm Optial Soiety of Ameria OCIS odes: ( ) Lasers, fiber; ( ) Fiber optis soures and detetors; ( ) Lasers, tunable; ( ) Optial oherene tomography. Referenes and links 1. D. Huang, E. A. Swanson, C. P. Lin, J. S. Shuman, W. G. Stinson, W. Chang, M. R. Hee, T. Flotte, K. Gregory, C. A. Puliafito, and J. Fujimoto, Optial oherene tomography, Siene 254(5035), (1991). 2. R. C. Youngquist, S. Carr, and D. E. N. Davies, Optial oherene-domain refletometry: a new optial evaluation tehnique, Opt. Lett. 12(3), (1987). 3. J. S. Park, M. Y. Jeong, C. H. Jung, C. H. Ouh, H. J. Kang, Y. G. Han, S. B. Lee, and C. S. Kim, Flexible urled optial ord for bending-insensitive optial imaging delivery, IEEE J. Sel. Top. Quantum Eletron. 16(4), (2010). 4. U. Sharma, N. M. Fried, and J. U. Kang, All-fiber ommon-path optial oherene tomography: sensitivity optimization and system analysis, IEEE J. Sel. Top. Quantum Eletron. 11(4), (2005). 5. J. S. Park, M. Y. Jeong, and C. S. Kim, Post-tuning of sample position in ommon-path swept soure optial oherene tomography, J. Opt. So. Korea 15, (2011), to be published. 6. A. B. Vakhtin, D. J. Kane, W. R. Wood, and K. A. Peterson, Common-path interferometer for frequenydomain optial oherene tomography, Appl. Opt. 42(34), (2003). 7. A. R. Tumlinson, J. K. Barton, B. Povazay, H. Sattman, A. Unterhuber, R. A. Leitgeb, and W. Drexler, Endosope-tip interferometer for ultrahigh resolution frequeny domain optial oherene tomography in mouse olon, Opt. Express 14(5), (2006). 8. J. A. Izatt, M. D. Kulkarni, H. W. Wang, K. Kobayashi, and M. V. Sivak, Optial oherene tomography and mirosopy in gastrointestinal tissues, IEEE J. Sel. Top. Quantum Eletron. 2(4), (1996). 9. S. W. Huang, A. D. Aguirre, R. A. Huber, D. C. Adler, and J. G. Fujimoto, Swept soure optial oherene mirosopy using a Fourier domain mode-loked laser, Opt. Express 15(10), (2007). 10. J. H. Lee, E. J. Jung, and C. S. Kim, Optial oherene tomography based on a ontinuous wave superontinuum seeded by erbium doped fiber s amplified spontaneous emission, J. Opt. So. Korea 14(1), (2010). 11. D. C. Adler, R. Huber, and J. G. Fujimoto, Phase-sensitive optial oherene tomography at up to 370,000 lines per seond using buffered Fourier domain mode-loked lasers, Opt. Lett. 32(6), (2007). 12. J. Zhang, B. Rao, L. Yu, and Z. Chen, High-dynami-range quantitative phase imaging with spetral domain phase mirosopy, Opt. Lett. 34(21), (2009). (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1102

2 13. M. A. Choma, A. K. Ellerbee, C. Yang, T. L. Creazzo, and J. A. Izatt, Spetral-domain phase mirosopy, Opt. Lett. 30(10), (2005). 14. C. Joo, T. Akkin, B. Cense, B. H. Park, and J. F. de Boer, Spetral-domain optial oherene phase mirosopy for quantitative phase-ontrast imaging, Opt. Lett. 30(16), (2005). 15. T. Endo, Y. Yasuno, S. Makita, M. Itoh, and T. Yatagai, Profilometry with line-field Fourier-domain interferometry, Opt. Express 13(3), (2005). 16. R. Huber, M. Wojtkowski, and J. G. Fujimoto, Fourier Domain Mode Loking (FDML): A new laser operating regime and appliations for optial oherene tomography, Opt. Express 14(8), (2006). 17. S. H. Yun and B. E. Bouma, Wavelength swept lasers, in Optial Coherene Tomography: Tehnology and Appliations, W. Drexler and J. G. Fujimoto, eds. (Springer, 2008). 18. S. H. Yun, G. J. Tearney, J. F. de Boer, and B. E. Bouma, Removing the depth-degeneray in optial frequeny domain imaging with frequeny shifting, Opt. Express 12(20), (2004). 19. D. Kim and Y. J. Cho, 3-Dsurfae profile measurement using an aousto opti tunable filter based spetral phase shifting tehnique, J. Opt. So. Korea. 12(4), (2008). 1. Introdution Optial oherene tomography (OCT) is useful for various sensing appliations, inluding medial imaging diagnostis as well as industrial surfae profilometry [1, 2]. A flexible optial probe is one of the key omponents of the OCT system, delivering the refleted optial signal from the sample to the imaging system at a remote loation. We have suggested a urled optial ord based on a holey fiber to be used as a bending-insensitive arbitrary-length optial imaging probe for a ommon-path OCT system [3]. However, there has been a ritial problem with this ommon-path OCT onfiguration in that the partial refletor needs to be plaed in between the sample and the objetive lens [4 7]. It is ritial to overome this limitation in order to ahieve the flexibility of sample positioning in front of the optial probe. There has been another problem of the fixed path-length differene in a ommon-path interferometer onfiguration, espeially for the phase-sensitive OCT appliations [8 15]. Sine the ommon-path interferometer uses the opropagation of referene and sample beams, it has the advantage of less phase noise ompared to the other interferometer onfigurations, and thus it is suitable for phase-ontrast profile imaging. However, the transverse resolution annot be redued below one mirometer beause it is diffiult to use an objetive lens with a high numerial aperture (NA) due to the limited path-length differene between the partial refletor and sample in a onventional ommon-path interferometer. In this paper, we report a novel double ommon-path interferometer onfiguration that enables the limitless tuning of the objetive distane of the sample loation. We add an extra dupliate ommon-path interferometer along the post-path of the interfered optial signal from the original ommon-path interferometer to generate an interferene signal that is independent of the optial distane between the partial refletor and sample. 2. Bakground 2.1 Limit of Mah-Zehnder interferometer for OCT Before we investigate the proposed onfiguration in detail, it is neessary to review the traditional OCT onfigurations and larify the limitations of various onventional optial interferometers. Figure 1(a) shows a shemati of the onventional OCT setup based on a Mah-Zehnder interferometer (MZI) and a urled optial fiber probe. When the path-length differene between the sample and referene mirror, l S l R, is less than the oherene position of light, the wavelength-swept light soure an indue the depth-resolved information of an OCT image. A detailed explanation of the experimental setup will be desribed later in Setion 3. Although the polarization dependene of two optial paths (a sample arm and a referene mirror arm) an affet the OCT image, it has not been a ritial issue when a onventional optial ord is used in both arms. However, when a urled optial ord is used in the sample arm as the flexible optial probe, we an learly see that the OCT image quality varies onsiderably, depending on the relative polarization state of the two separate arms. Figure 1(b) shows two examples of the OCT image on a fish eye sample generated by tuning the polarization ontrollers of the two separate arms (one good quality and one poor quality). As the urled optial ord is hanging down freely between the sample and imaging system, (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1103

3 the amplitude of the interferene signals beomes extremely sensitive to the shape and loation of the urled optial ord. Thus, the quality of the OCT image also varies rapidly in real time between good and poor due to the variation in the dynami range of the signal. Sine the logarithm intensity signal of Fig. 1(b) is normalized to the highest peak signal level in eah A-san, the deayed dynami range is displayed with a vertial strip noise pattern due the enhaned noise level relatively. Otherwise, if the logarithm intensity signal is normalized to the noise signal level, the deayed dynami range will be displayed with the redued peak signal in eah A-san. In onlusion of this setion, we simply showed the problems in using MZI for a flexible optial probe OCT. (a) (b) Light Soure Detetor 50:50 Coupler Polarization Controller Curled Optial Fiber Objetive Lens Sample l S - l R Referene Mirror Fig. 1. (a) Shemati of the onventional OCT setup based on a Mah-Zehnder interferometer and a urled optial fiber probe. (b) Variation of OCT image quality depending on the polarization state by tuning polarization ontrollers. 2.2 Limit of onventional ommon-path interferometer for OCT Beause the MZI is not suitable for the flexible optial probe of OCT, a ommon-path interferometer has been developed to overome the polarization mismath between the referene mirror arm and sample arm of the urled optial ord [3]. Sine there is only one path in the ommon-path interferometer, as shown in Fig. 2(a), the refleted signals from the referene mirror and sample are propagating separately through a ommonly shared optial path [4 7]. However, there is still a ritial problem with the onventional ommon-path OCT in that the partial refletor must be loated inonveniently lose to the sample beause the interfered optial signal an be produed only when the delay between the sample arm, l S, and the referene mirror arm, l R, is less than the oherene position. Figure 2(b) illustrates an experimental example of this limitation. When the objetive distane between the partial refletor and sample, l S l R, is 1 mm, the amplitude of the interferene signal is strong enough to obtain the OCT image learly. As we inrease the objetive distane to 2 mm and 3 mm by moving the partial refletor position, it is lear that the amplitude of the interferene signal is dereased and the quality of the OCT image is proportionally degraded due to the inreased sampling frequeny and deayed dynami range of the signal. This result indiates that the flexibility of a urled bending, an arbitrary probe length, and a polarization math an be ahieved suessfully due to the ommon-path interferometer; however, the flexibility of the sample position in an optial probe is still a problem that needs to be solved. In onlusion of this setion, the onventional ommon-path interferometer has also the problems for a flexible (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1104

4 optial probe OCT. It is ritial that we find a method that will enable us to get a good-quality image without requiring us to position the sample lose to the partial refletor. (a) Light Soure Detetor Cirulator Curled Optial Fiber Objetive Lens Partial Refletor Sample l R l S - l R l S (b) l S l R = 1 mm l S l R = 2 mm l S l R = 3 mm 4.5 mm 3.0 mm Fig. 2. (a) Shemati of the onventional OCT setup based on a onventional ommon-path interferometer and a urled optial fiber probe. (b) Variation of OCT image quality depending on the objetive distane by moving the position of the partial refletor. 3. Experimental setup 3.1 Setup for double ommon-path interferometer Sine neither the MZI nor the onventional ommon-path interferometer is suitable for the flexible optial probe of OCT, we propose a novel onfiguration for a double ommon-path interferometer, as shown in Fig. 3(a). To generate an interferene signal regardless of the optial distane between the partial refletor and sample, an extra dupliate ommon-path interferometer an be added along the post-path of the interfered optial signal from the original ommon-path interferometer. The output of the wavelength-swept laser passes through irulator 1, ommon-path 1, irulator 2 and ommon-path 2, in series, and the photodetetor. Light from the seond port of irulator 1 is partially refleted in order to generate a referene signal (E R ) from the partial refletor 1. The rest of the transmitted light is foused on the sample in order to generate a sample signal (E S ) from eah surfae of the sample. Assuming a single refletion in the sample, the eletri field (E 1 ) for ommon-path 1 is represented by E E E 1 R S j2 nf lr j2 ( nflr nad1 ) R RSe R e where ω (= 2πν) denotes the temporal angular frequeny of the light soure, n F and n A denote the refrative indexes of orresponding media of mainly optial fiber and air, respetively, R R and R S denote the refletion oeffiients of partial refletor 1 and the sample, respetively, n A d 1 denotes optial path-length differene between the partial refletor and sample, and n F l R denotes the ommon optial distane from irulator 1 to the partial refletor 1. (1) (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1105

5 Fig. 3. Shemati of the proposed double ommon-path OCT setup. Common-path 2 indues an additional path-length differene of 2n A d 2 to aquire the interferene signal in any ase in whih the objetive distane between the partial refletor and sample, 2n A (l S l R ) or 2n A d 1, is larger than the oherene position, l C. The signal E 1 from ommon-path 1 enters irulator 2 and ommon-path 2, in series, and is then refleted by partial refletor 2 and the mirror in ommon-path 2. Sine E R and E S are refleted together, four terms (E R1, E S1, E R2, and E S2 ) are generated in ommon-path 2. We an represent the eletri field (E 2 ) for ommon-path 2 as E E E E E 2 R1 S1 R2 S 2 j2 ( nf lr nflp ) j 2 ( nflr nflp nad1 ) R P RS RP e R R e j2 ( nflr nflp nad2 ) j 2 ( nflr nflp nad1 nad2 ) R M RS RM e R R e where R P and R M denote the refletion oeffiients of partial refletor 2 and the mirror, respetively, and n F l P denotes the ommon optial distane from irulator 2 to partial refletor 2. In the ase of tuning the relative path-length differene to 2 na( d1d2) lc, the E S1 and E R2 terms in Eq. (2) an be used to generate an interferene fringe. The deteted photourrent I an be expressed as * I E E E E E E E E (3) R1 S1 R2 S 2 R1 S1 R2 S 2 where ρ is the responsivity of the detetor, and the angular braket denotes integration over the response time of the detetor. In this ase, the interferene term of the deteted photourrent I an be expressed as I RR RS RP RM os 2 na( d1d2) (4) This theoretial desription implies that an optial distane that is larger than 2n A d 1 between a sample and partial refletor 1 in ommon-path 1 an generate interferene suessfully by adjusting the extra length of 2n A d 2 in ommon-path 2. For the generation of an interfered optial signal for the OCT image, we applied a sweptsoure OCT onfiguration by using an FDML swept laser with a 2.8 km delayed fiber in the ring avity to operate at a repetition rate of khz [16]. The tuning range was 102 nm entered at a wavelength of 1295 nm, and the output power was 9 mw. The oherene position (l C ) was around 14.7 mm, whih was estimated from the spetral linewidth of 0.05 nm for the laser [17]. The laser output was not filtered or amplified to ensure phase stability. A total of 5% of the laser output propagated to the MZI for resampling the aquired data by interpolating to a uniform spaing in k-spae. The MZI generated a omb-like signal (2) (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1106

6 orresponding to a frequeny shift of GHz. The realibration signal was reorded on the seond hannel of an A/D onverter. The interferene signal reorded from the first hannel was realibrated to a frequeny shift of the MZI to ompensate for the nonlinear frequeny sweep of our swept laser. It was then interpolated and fast Fourier transformed to aquire exat phase values [18]. For ahieving improved stability for the OCT imaging, same kind of an objetive lens and a partial refletor were used in ommon-paths 1 and 2 beause the similar optial beam ondition of the objetive lens and partial refletor between the two ommon-paths will redue the mismath of length, polarization, and dispersion ondition of the interfered optial signal [5]. To redue environmental vibration, the partial refletor and sample in ommonpath 1 were mounted on one stage. The partial refletor and mirror in ommon-path 2 were also mounted on one stage. The partial refletors in ommon-paths 1 and 2 onsist of a over glass with two sides. One side, with an antirefletion oating, had a refletivity below 0.002, and the other side had a normal refletivity of 0.05 in the 1300 ± 50 nm region. For the optimal interferene signal, the refletivity of the partial refletor and mirror ould be tuned by an additional surfae oating. Sine the redued sensitivity an be aused by the addition of a partial refletor in ommon-path 2, the analysis of sensitivity is required to provide a quantitative optimization using Eq. (4). Considering the refletivity of mirror, R M is 1 in ommon-path 2, we an obtain the optimally enhaned sensitivity when the refletivity of the partial refletor R P in ommon-path 2 is around 0.33 [4]. By using an additional oated refletor, we expet to enhane the sensitivity of the present method The loation of the partial refletor along the foal beam also helped to ontrol the refleted signal intensity from the partial refletor [5]. Depending on the appliation of OCT, it is sometimes neessary to insert an additional objetive lens between the partial refletor and sample in the ommon-path OCT system. This an be implemented suessfully by using the proposed double ommon-path interferometer. For example, the length of a ommerial objetive lens body in our laboratory was measured to be between 28 mm (the lowest NA of 0.1) and 41 mm (the highest NA of 0.85), and thus the path-length differene of 2n A d 1 between E R and E S is always more than l C (~14.7 mm) of the wavelength-swept laser light in our OCT system. Therefore, an interferene fringe annot be observed in general when deteting the optial signal of E 1 from ommon-path 1. To deal with this ase, an interferometer with an extra ommon-path 2 an be useful. This an be ahieved by preparing two idential objetive lenses and partial refletors to adjust the pathlength differene without limitation, as shown in Fig Phase-sensitive OCT probe with high NA objetive lens To illustrate the usefulness of an arbitrary distane between the sample and partial refletor due to the proposed double ommon-path interferometer, we demonstrate a phase-sensitive OCT experiment with an enhaned performane ahieved by using an objetive lens with a high NA over This phase-sensitive OCT tehnology [11] is also known as Spetral Domain Phase Mirosopy (SDPM) [12, 13] or Spetral Domain Optial Coherene Phase Mirosopy (SD-OCPM) [14]. In order to improve the axial resolutions of OCT below the typial resolution level of a few mirometers, whih results from the onventional intensity information of the interferene signal, the phase term from the interferene signal an be used to measure the sub-oherene-length variations in the axial dimension of the sample. Reently, phase-sensitive OCT has been demonstrated to ahieve sub-mirometer axial resolution using the ommon-path interferometer onfiguration [11]. Sine phase stability is ritial for measuring the variation in interferene, the ommon-path interferometer is preferred over other interferometers with separate beam paths, suh as the MZI and Mihelson interferometer [11 14]. In ontrast, the transverse resolution annot be redued below one mirometer beause it is diffiult to use a high-na objetive lens for the onventional ommon-path interferometer [11 14]. Sine, in general, a thin over glass (thikness 160 μm) should be inserted in the short working distane between the objetive lens and sample to serve as a referene refletor, (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1107

7 sub-mirometer lateral resolution has not been demonstrated with the ommon-path onfiguration beause there is a limited spae between the sample and high-na objetive lens to insert a over glass [11 14]. However, in the ase of the double ommon-path onfiguration, a high-na objetive lens an be positioned near the ontat loation of the sample suessfully by moving the partial refletor far from the sample, as shown in Fig. 4. For improved transverse resolution in phase-sensitive OCT, we used two idential objetive lenses (NA = 0.85) for both ommon-path interferometers. These 0.85-NA objetive lenses are the ones with the highest NA in our laboratory, and the foal length is as short as 2.9 mm (working distane = 300 μm; foal depth = 0.4 μm). The theoretial transverse resolution is 1.5 μm for an objetive lens with an NA of 0.4, but this an be improved up to 0.7 μm for an objetive lens with a high NA of If we use a single ommon-path interferometer by plaing a over glass with a thikness of 160 μm within a working distane of 300 μm, there may be surfae damage on the lens or sample and limited ontrol of the refleted signal beam from the partial refletor. Common-path 1 Partial Refletor 160 mm Working distane = 300 mm Sample High NA: 0.85 (60x) Common-path 2 Foal depth = 0.4 mm Objetive Lens Body= 41 mm Mirror High NA: 0.85 (60x) Fig. 4. Shemati of a phase-sensitive OCT probe based on the double ommon-path interferometer setup, as shown in Fig Experimental results Based on the proposed onfiguration of double ommon-path OCT, we performed two experiments. The first is the deep OCT imaging independent of the optial distane between the partial refletor and sample using a low-na objetive lens, and the seond is the improved transverse-resolution phase-sensitive OCT using a high-na objetive lens. 4.1 Low-NA objetive lens probe for deep OCT imaging Figure 5 shows the ross-setional images of a fish eye (Balantioheilos melanopterus) aquired at different objetive distanes between the partial refletor and sample using a double ommon-path OCT system. The results are ompared with those in the ase of the previous OCT imaging obtained with the same sample using single ommon-path OCT, as shown in Fig. 2, where the image quality is degraded due to the dynami range roll-off as the objetive distane inreases. In the measurement of Fig. 5, the objetive distane between the partial refletor and sample, l S l R or d 1, is inreased to 6 mm and 16 mm in ommon-path 1 only that results in degradation of quality and disappearane of the OCT image, respetively. As we also inrease the extra delay distane, d 2, in ommon-path 2 to move the peak for the point spread funtion (PSF) toward the 0 mm position of the optial path-length differene, the OCT image an be restored to its original quality, where the sample is near the partial refletor. We an see that the quality of both Figs. 5(a) and 5(b) is similar to that of Fig. 2(b), where l S l R = 1 mm. The images onsist of 360 axial sans (transverse pixels) with 630 (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1108

8 points (axial pixels) per san. The idential objetive lenses used in the measurements of Fig. 2 and Fig. 5 have a low NA of 0.1 for deep OCT imaging. It seems that the dynami range of double ommon-path OCT is less than that of Mah-Zehnder- or Mihelson-type OCT due to the extra loss term in Eq. (2); however, this onfiguration has the advantage of enabling a flexible OCT image based on a urled optial ord, as proven by a omparison of the experiment results of Fig. 1 and Fig. 2. Although the extra delay distane, d 2, may be implemented by other interferometers of Mah-Zehnder or Mihelson type instead of ommon-path 2, the ommon-path interferometer with an idential objetive lens has the advantage of enhaned environmental stability [5]. (a) l S l R = 6 mm (b) l S l R = 16 mm 4.5 mm 3.0 mm Fig. 5. Maintaining OCT image quality for different objetive distanes of (a) 6 mm and (b) 16 mm, respetively, using the proposed double ommon-path OCT system with a low NA objetive lens (NA = 0.1). 4.2 High-NA objetive lens probe for phase-sensitive OCT The harateristis of a high-na objetive lens, inluding high transverse resolution, short working distane, and shallow foal depth, enable us to obtain a phase-ontrast profile image with a fine struture, whih is a good appliation of phase-sensitive OCT [19]. Before showing the profile measurement for the proposed phase-sensitive OCT onfiguration, we ompare the effet of the transverse resolution of objetive lenses with different NA values by using two en-fae optial oherene mirosopy (OCM) images of the USAF 1951 resolution target (group 7), as shown in Figs. 6(a) and 6(b). We seleted a small imaging area of group 7 to demonstrate a marginal omparison of transverse resolutions of different NA values. An objetive lens with a low NA of 0.4 is used to generate the en-fae OCM image with a onventional ommon-path OCT system shown in Fig. 2 beause it has a relatively long working distane between the sample and objetive lens to insert the partial refletor. In omparison, using the double ommon-path OCT onfiguration shown in Fig. 4, two idential objetive lenses with an NA of 0.85 are also used to generate the en-fae OCM image. Sine the theoretial transverse resolution is 1.5 μm and 0.7 μm for the objetive lenses with an NA of 0.4 and 0.85, respetively, the experimental results learly show the differene in quality between these two onfigurations. Figure 6(a) illustrates that a onventional phase-sensitive OCT with of 0.4 NA an marginally distinguish the spaing in element 2 with a somewhat blurred image. In ontrast, we an learly resolve the smallest spaing in element 6 using a double ommon-path onfiguration with an objetive lens of 0.85 NA, as shown in Fig. 6(b). It is noted that the linewidths of elements 2 and 6 are 3.10 μm and 2.19 μm, respetively. Two axis-translation stages are used to move the sample along the transverse axis under the fixed position of the objetive lens in the sample arm. The slight distortion in the OCM image an be improved by tuning the path angle of the translation stage. Though the high NA lens with a short working distane is less onvenient to san along transverse diretion, the removal of partial refletor between lens and sample an be an advantage of the proposed double ommon-path onfiguration. A pair of Galvano mirrors an replae the translation stage to enhane the imaging speed. (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1109

9 In Figs. 6() and 6(d), the phase stability is ompared for both the single and double ommon-path onfigurations, respetively. For the same light soure as for the FDML swept laser, we measured the phase variations in both onfigurations during 500 axial sans for a mirror sample in a stati position. The measured phase variations were onverted into time traes of displaement variations. The displaement variations were omputed using the relation z 0 /4n, where 0 is the enter wavelength of the light soure, and n is the average group index between the partial refletor and mirror sample refletors [12]. After measuring from both onfigurations during 500 axial sans, the displaement variations were determined. The standard deviations were pm and pm for the onventional and proposed double ommon-path onfigurations, respetively. It appears that the inreased phase noise is the result of extra environmental instability and aberration assoiated with the addition of ommon-path 2. We expet that the additional instability an be redued by a shorter length for extra opti omponents and the replaement of the irulator with a diretional oupler. Nevertheless, it is a more ritial advantage that the double ommon-path interferometer an enhane the transverse resolution due to the use of a higher-na objetive lens and an support the flexibility of the sample position from the objetive lens. Even with the inreased displaement variation, the system an measure a nanometer-sale axialdisplaement profile. Fig. 6. Conventional phase-sensitive OCT (NA = 0.4) measures (a) en-fae OCM image and () time trae of displaement; double ommon-path phase-sensitive OCT (NA = 0.85) measures (b) en-fae OCM image and (d) time trae of displaement. Figure 7 shows nanometer-sale 3D phase-ontrast profile images that were obtained using a double ommon-path phase-sensitive OCT onfiguration with an objetive lens with a high NA of For a sample of the USAF 1951 resolution target, the 3D OCT data set, onsisting of voxels, is reorded for both the en-fae OCM images and the phase-profile images. The intensity profile (Fig. 6(b)) and phase profile (Figs. 7(a) and 7(b)) are imaged from one data set measurement. The measurement of the phase variation orresponds to the sub-oherene-length variation in the seleted sample layer given by the simple relation z 0 /4n, where 0 is the enter wavelength of the light soure, is the unwrapped phase variation, and n is the refrative index between the partial refletor and sample [12]. The onverted height profile of the patterned hromium layer has a (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1110

10 nanometer-sale axial resolution. By using a high-na objetive lens, the spaings of the smallest elements an also be learly resolved with sub-mirometer transverse resolution. 180 mm (a) 40 mm (b) nm (a) mm 0 80 mm 100 mm () (d) nm mm Fig. 7. 3D phase-ontrast surfae profile images of (a) group 7, (b) enlargement of elements 5, 6 of the USAF 1951 resolution target, () a hologram pattern with μm 2 area, and (d) enlargement of the marked region (20 20 μm 2 ). To demonstrate potential industry appliations, we repeated the experiment using a hologram pattern on a photoresist substrate. A positive photoresist substrate (AZ 151) was exposed to UV light in order to reate holes that were 2 μm in diameter and 200 nm in depth. This sample was an intermediate produt before eletroforming. From the 3D surfae profile images of Figs. 7() and 7(d), small variations within the sub-oherene length, suh as hole depth, one shape symmetry, and surfae debris defets, an be learly monitored with high axial and transverse resolutions. Figure 7() was imaged from a 3D data set of voxels. The entire 3D data were aquired by reording eah 2D data set in sequene on a omputer system memory. This 3D imaging information is useful to hek for a math with the designed parameters and to provide feedbak to optimize the fabriation proess in real time. Using the phase-sensitive OCT system, it was impossible to measure the same sample with a similar axial and lateral resolution simultaneously. When the same sample was measured by atomi fore mirosopy (AFM), it was diffiult to get an exat profile beause the sample surfae was damaged by the physial ontat of the AFM probe tip with the target. 5. Conlusion In summary, we demonstrated a novel OCT system based on a double ommon-path interferometer that enables flexible urled bending, arbitrary probe length, and probe-sample separation. Our proposed onfiguration enhaned the onveniene of sample position in ommon-path interferometer, like the other Mihelson and Mah-Zehnder interferometers. By adjusting the additional delay time by the seond ommon-path interferometer, the optial path-length differene of sample position an be moved to near the zero-delay position where high dynami range is obtained by lowering the sampling frequeny. In addition, with the allfiber double ommon-path onfiguration, the phase noise an be redued so that our system an be applied to the phase-sensitive OCT imaging with high transverse resolution. We obtained simultaneously a high axial-resolution imaging with nanometer-sale phase stability and a high transverse-resolution imaging with sub-mirometer resolving power using a phasesensitive OCT onfiguration with two idential high-na objetive lenses. We expet that the (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1111

11 3D surfae profile image will be more pratial for the in-line produt inspetion of finestruture samples and for in vivo endosopi imaging of biomedial samples. Aknowledgments This researh was supported by the World Class University program through the National Researh Foundation of Korea funded by the Ministry of Eduation, Siene and Tehnology, South Korea (grant no. R ). This work was also supported by the National Institute of Health, USA (EB-10090, HL , RR-01192) and by the Tehnology Innovation Program ( ) funded by the Ministry of Knowledge Eonomy, Korea. (C) 2012 OSA 16 January 2012 / Vol. 20, No. 2 / OPTICS EXPRESS 1112