Open Access
Add to BibSonomyAbstract
We present a high-speed wavelength-swept fiber laser based on a dispersion tuning method using a reflective semiconductor optical amplifier (RSOA) and a chirped fiber Bragg grating (CFBG). By using these devices, the cavity length can be shortened drastically. The short cavity improves the laser performance at high sweep rates over 200kHz. We achieve a sweep range of 60nm and an output power of 8.4mW at 100kHz sweep. We applied the dispersion-tuned fiber laser to the swept-source OCT system and successfully obtained OCT images of an adhesive tape at up to 250kHz sweep rate.
©2013 Optical Society of America
1. Introduction
Optical coherence tomography (OCT) is a noninvasive cross-sectional imaging technique using infrared light, which can be used to obtain images of tissue with micron scale resolution [1]. Especially, swept-source OCT (SS-OCT) using a wavelength-swept laser has attracted a lot of attentions in recent years due to its large imaging depth and high imaging speed [2, 3]. In the SS-OCT system, the performance of the wavelength-swept laser dominate the system performance.
Up to now, a number of broadband swept light sources for SS-OCT have been reported [4–9]. However, most of the conventional swept light sources have a fundamental limitation in the sweep speed caused by the tuning speed of the filter and the photon lifetime in the laser cavity. To solve the latter problem, a method called Fourier domain mode-locking (FDML) was proposed by R.Huber et al. [10]. The FDML laser includes a long dispersion-managed delay fiber in the ring cavity, and an entire frequency sweep is stored within the delay line. Therefore, the FDML laser is not limited by the photon lifetime. Using this method, the sweep rate of the laser has been improved significantly [11–14].
We have demonstrated a wavelength-swept fiber laser based on a technique called dispersion tuning without the need for the filter [15–18], which can solve the former problem, i.e., the limitation caused by the tuning speed of the filter. We have reported a wavelength-swept laser which has a tuning range of over 100nm with a sweep rate of 200kHz [15–17]. The dispersion-tuned fiber laser at 1.3µm bands has been applied to the SS-OCT system and an OCT image of the human finger was successfully obtained at 1kHz sweep rate [18]. However, we could not obtain OCT images at higher sweep rates because of the performance degradation of the laser. Since we used 100m-long dispersion compensating fiber (DCF) as the dispersive element needed for dispersion tuning, the total length of the cavity exceeded 100m. The long cavity increased the photon lifetime, and resulted in a decrease in the output power and the linewidth broadening at higher sweep rates. In order to enhance the laser performance at higher sweep rates, the cavity length has to be shortened.
In this paper, we use a chirped fiber Bragg grating (CFBG) and a reflective semiconductor optical amplifier (RSOA) to shorten the cavity length of the dispersion-tuned fiber laser. The dispersion-tuned laser using the CFBG as the dispersive element was first demonstrated by Burgoyne et. al. [19–21]. We have replaced the DCF with the CFBG and achieved a sweep rate of 500kHz [22]. The RSOA has been applied to dispersion-tuned fiber laser to enhance the modulation and dispersion efficiency [23]. We combine the CFBG and the RSOA to make a linear cavity instead of a ring cavity that tends to be longer. The cavity length is shortened to about 4.1m and the performance at high sweep rate over 200kHz is significantly improved. The tuning range of the laser is about 60nm and the average output power is 8.4mW. We successfully obtain OCT images at sweep rates up to 250kHz using this laser. We compare the performance of three types of dispersion-tuned fiber lasers, i.e., the laser using the DCF, the one using the CFBG, and the one using the RSOA and CFBG, and demonstrate that the last one predominates over the other ones.
2. Principle
2.1 Dispersion tuning
The principle of the dispersion tuning method is mode locking of a dispersive laser cavity. As shown in Fig. 1(a), a highly dispersive element (e.g., DCF) is inserted into the laser cavity and the light in the cavity is intensity modulated for active mode locking. The lasing wavelength changes corresponding to the frequency of intensity modulation.
The tuning process is explained as follows [15–18]. When the light in the laser cavity is intensity modulated at the integral multiple frequency of the free spectral range (FSR), the light is actively mode-locked. The FSR of the laser cavity F can be expressed as;
where L is the cavity length, n is the refractive index in the cavity, and c is the speed of light in vacuum. When the cavity contains chromatic dispersion, the FSR is a function of wavelength or frequency f. That is, when we apply a modulation at fm to the dispersive cavity, the laser is forced to operate at the wavelength to meet the harmonic mode-locking condition. This tuning process is depicted in Fig. 1(b). The change of lasing wavelength can be expressed as;where D is the dispersion parameter in ps/nm/km, Dtotal is the amount of dispersion in ps/nm, fm0 is the center of the modulation frequency, n0 is the refractive index at fm0, L is the cavity length, andis the change of the modulation frequency. This equation indicates that the lasing wavelength can be swept linearly by sweeping the modulation frequency. The wavelength tuning range is determined by the gain bandwidth of the gain medium and the lasing wavelength of adjacent harmonic modes. The maximum tuning rangecan be achieved when the change of modulation frequencyexceeds the FSR. From Eq. (2), can be expressed as;where F0 is the FSR at fm0. This equation indicates that smaller fm0 and Dtotal are needed for wider tuning ranges. However, smaller fm0 increases the instability of lasing wavelength and smaller Dtotal means the difference of the FSR between wavelengths is smaller, and both these facts lead to linewidth broadening. Therefore, we have to optimize the parameters fm0 and Dtotal.2.2 OCT performance
The performance of the OCT system is determined by resolution, depth range, imaging speed and sensitivity. For a tuning source with a Gaussian-profile spectral envelope, the axial resolution of a OCT system can be expressed as [24]
where n is the refractive index of the sample,is the center wavelength andis the full width at half maximum (FWHM) of the spectral envelope (tuning range). As Eq. (4) indicates, the axial resolution is inversely proportional to the sweep range. Therefore, a wavelength-swept laser with wide tuning range is needed.The coherence length of the SS-OCT system is defined as twice the depth at which the signal intensity drops by 6dB. The relation between the coherence length lc and the instantaneous linewidth of the laser can be expressed as [25];
which indicates that a narrower instantaneous linewidth is necessary for a deeper imaging range.The imaging speed of the SS-OCT system is proportional to the sweep rate of the laser. It also depends on the number of pixels needed for images. For applications such as endoscopy and vascular catheters, short inspection time is crucial, therefore wavelength-swept lasers with over 100kHz sweep rate are in strong demand. The sensitivity of the system is the signal-noise ratio of the OCT signal, a laser with high power and low noise is needed for high sensitivity.
3. Laser setup and characteristics
3.1 Experimental setup
Here, we compare the performance of three types of dispersion-tuned fiber lasers shown in Fig. 2. The first setup using DCF is shown in Fig. 2(a). The laser consists of a ring resonator, including a semiconductor optical amplifier (SOA) module (SOA1013, COVEGA), a 100m-long DCF (−90ps/nm/km at 1.5µm bands), a polarization controller (P.C.) and an isolator. 10% of the intracavity light is output via a fiber coupler. The laser output is amplified by the second SOA (BOA1004, COVEGA). The cavity length is 105.2m.
The second configuration shown in Fig. 2(b) is the same setup as the first one except that the CFBG (−10ps/nm at 1.5µm bands, TeraXion) is used with a circulator instead of the DCF. The CFBG is made in a PM fiber and with a reflectivity of 65%. The cavity length is 9.8m.
By combining the RSOA and the CFBG, we can adopt a linear resonator configuration, as shown in Fig. 2(c). The light from the RSOA (IPSAD1501, INPHENIX) enters the CFBG and 65% of the light is reflected. The remaining 35% of the light passes through the CFBG and is amplified by the SOA. Since all the fibers used in this configuration are PM fibers, a P.C. is not needed in the cavity. The cavity length is 2.05m. Since this is a linear laser, the effective length is doubled to be 4.1m, less than half of that in the ring laser.
Direct current (DC) for driving SOA and alternate current (AC) for intensity modulation are injected into the SOA through the bias tee. A synthesizer is used for generating the AC mode-locking signal. As the cavity includes a highly dispersive element, the lasing wavelength changes corresponding to the modulation frequency. This can be confirmed by changing the modulation frequency manually. In order to sweep the laser wavelength, a saw-tooth wave is input to the synthesizer from the function generator (FG) to sweep the modulation frequency linearly. The output spectra and the temporal waveform are observed by an optical spectrum analyzer and an oscilloscope, respectively.
3.2 Laser characteristics
We set the center modulation frequency fm0 at around 411MHz, 612MHz and 770MHz for the setup-A (SOA + DCF), the setup-B (SOA + CFBG) and the setup-C (RSOA + CFBG), respectively. As we mentioned in section 2.1, there is a trade off in the choice of fm0: the lower fm0 leads to linewidth broadening, and the higher fm0 to limited tuning range. We first measured the responses of the SOA modules as a function of modulation frequency to find the best fm0. The fm0 chosen in each setup is the highest modulation frequency at which the maximum tuning range is maintained. The drive current to the SOA1 in setup-A and setup-B is 110mA and that to the RSOA in setup-C is 55mA. The drive current to the SOA2 is around 300mA. The AC modulation signal from the synthesizer is amplified to 28dBm and input to the SOA via the bias tee.
Figure 3 shows the static tuning characteristics of the dispersion-tuned wavelength-swept fiber lasers. Figures 3(a)-3(c) shows the lasing spectra of setup-A, setup-B and setup-C observed by the OSA. The tuning range of each setup are 131nm, 77nm and 83nm. The tuning range of the setup using CFBG are narrower than that using DCF. This is because the reflection bandwidth (FWHM) of the CFBG is 100nm and the tuning range of the laser is limited to less than 100nm. The tuning range around 80nm correspond to the axial resolution of 13µm using Eq. (4). Since the dispersion-tuned laser does not have the Gaussian spectral envelope, the actual resolution would be lower. Figures 3(d)-3(f) shows the relation of the lasing wavelength and the mode-locking frequency. We found that the lasing wavelength is linearly proportional to the mode-locking frequency.
Fig. 3 Static characteristics. (a)-(c) The lasing spectra of setup-A, B, C. (d)-(f) The relation of the lasing wavelength and the mode-locking frequency in setup-A, B, C.
In the dispersion tuning method, the change of the optical frequency, not wavelength, is proportional to the mode-locking frequency when the higher-order dispersions () are negligible [17]. Thus the dispersion-tuned laser having only 1st-order dispersion is suitable for the SS-OCT system, which requires the linear frequency sweep for the linear sampling in the optical frequency domain. The average static linewidths of the setup-A, setup-B and setup-C are 0.59nm, 0.45nm and 0.32nm, respectively. The linewidth becomes narrower as the cavity length become shorter. We presume this is due to the use of higher mode-locking frequency for shorter cavity.
Figure 4 shows the sweep characteristics obtained by inputting the saw-tooth wave from the FG and sweeping the mode-locking frequency linearly. Figures 4(a)-4(c) shows the peak-hold spectra at the sweep rate of 1kHz, 10kHz, 100kHz and 1MHz. The decrease in the optical spectra at high sweep rates is because of the integral time of the peak-hold function of the OSA. The ripples observed in Figs. are presumably due to the modulation characteristics of SOA and RSOA. The sweep range of the setup-A is intentionally reduced to 80nm in order to compare the sweep characteristics with the other two setups under the same spatial resolution. The peak-hold spectrum at 1MHz of the setup-A collapses due to its long cavity length. The sweep range of the setup-C shown in Fig. 4(c) is about 60nm, which is narrower than the tuning range shown in Fig. 3(c). We had to limit the sweep range to 60nm. The sweep of modulation frequencyneeded for a specific sweep range is inversely proportional to the cavity length as indicated in Eq. (2), and our RF synthesizer and FG are not capable of covering the entire 83nm range. The output power of setup-A, setup-B and setup-C are 9.5mW, 4.9mW and 8.4mW, respectively.
Fig. 4 Dynamic characteristics. (a)-(c) The peak-hold spectra of setup-A, B, C. (d)-(f) The temporal waveform of setup-A, B, C.
Figures 4(d)-4(f) shows the temporal waveforms at 1kHz, 10kHz, 100kHz and 1MHz. The voltage of setup-A output shown in Fig. 4(d) decreases as the sweep rate becomes faster, and no signal is observed at the sweep rate of 1MHz. The waveforms of setup-B shown in Fig. 4(e) is superior to that of setup-A. However, the signal could hardly be observed at 1MHz. The sweep waveforms of setup-C are much better than the others. As shown in Fig. 4(f), the output power of setup-C remains high at fast sweep speed of 1MHz. This indicates that this laser can work at high sweep rate exceeding 100kHz. The performances of each setup is summarized in Table 1. The static linewidth and the laser performance at high sweep rates become better as the laser becomes shorter.
Table 1. Summary of Performances of Three Lasers
4. OCT application
We applied the dispersion-tuned wavelength-swept fiber laser to the OCT system. We adopt the INNER VISION OCT system from Santec Corp [26]. Figure 5 shows the setup of the SS-OCT system. We replaced the light source of the system with our dispersion-tuned laser.
Figure 6 shows the point spread function (PSF) of the laser at the sweep rate of 1kHz and 50kHz. The PSF is measured with a Michelson interferometer in Fig. 5 by placing and moving a 90% mirror at the probe arm. The amplitude of the interference fringes decreases as the optical path difference is longer according to the coherence length, which is equal to the imaging depth. Figure 6(a)-6(c) shows the PSF of setup-A, setup-B and setup-C at 1kHz sweep rate. The axial resolution of the systems read from the FWHM of the signal in PSF are about 30µm, which are nearly the same for all setups. As the theoretical value calculated from Eq. (4) is about 10µm, the axial resolution of the OCT system is three times worse than the theoretical value. We attribute it to the mismatch between the laser and the adopted OCT system, i.e., the laser is pulsed whereas the OCT system is optimized for continuous-wave laser operation. Figures 6(a)-6(c) indicates that the depth range of setup-A is worse than the other two setups. The coherence length of each system calculated from Eq. (5) are 1.7mm, 2.2mm, 3.1mm, and measured from the PSF are 0.6mm, 2.2mm, 1.6mm. The setup using the normal SOA and the CFBG shows the best drop-off characteristics. The measured coherence length of other two setup are worse than calculated value, which indicates the linewidth broadening in wavelength sweep. In setup-A, it is because of a long cavity length. In setup-C, the linewidth is probably not constant through the sweep due to the modulation efficiency of the RSOA.
Fig. 6 Point spread function (a)-(c) at 1kHz ((a) setup-A (b) setup-B (c) setup-C). (d)-(f) at 50kHz ((d) setup-A (e) setup-B (f) setup-C).
Figures 6(d)-6(f) shows the PSF of setup-A, setup-B and setup-C at 50kHz sweep rate. As shown in Fig. 6(d), the signal drop-off of setup-A is not good and we could observe the signal only near the surface. The performance of setup-B shown in Fig. 6(e) is better than that of setup-A. Still, the coherence length of setup-B is about 0.5mm, which is worse than that of 1kHz sweep. By contrast, the drop-off characteristics of setup-C are better, as indicated in Fig. 6(f). The coherence length is about 0.8mm, which is better than that of the other two setups.
Figure 7 shows the OCT images of an adhesive tape obtained by the system. In each image, several layers from the surface are observed. As shown in Fig. 7, we could obtain the image at the sweep rate of up to 10kHz using setup-A. At 50kHz, however, we could observe only the surface and could not penetrate inside the tape. We could obtain the image at up to 250kHz sweep rate using setup-B, although the images obtained at 125kHz and 250kHz are poor. In contrast, images from setup-C remained relatively good even at high sweep rates. The short cavity enhanced the stability at high sweep rate.
5. Conclusion
We improved a wavelength-swept fiber laser based on the dispersion tuning method using a RSOA and a CFBG. We demonstrated that by using a short linear cavity laser the performance at high sweep rates can be significantly enhanced. Using this laser, we successfully obtained OCT images of an adhesive tape at up to 250kHz sweep.
Although the sweep rate of the dispersion-tuned fiber laser improved, the coherence length of the OCT system is not sufficient due to linewidth broadening. One solution to this problem is using a dispersive element with anomalous dispersion. The combination of anomalous dispersion and self-phase modulation in SOA have been demonstrated to be able to narrow the linewidth of the laser [19–21, 27].
Acknowledgments
The authors wish to thank Mr. Masataka Tei of Santec Corp. for cooperation on the OCT experiment. This work was supported by the Funding Program for Next Generation World-Leading Researchers (NEXT Program) of the Japan Society for the Promotion of Science (JSPS).
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(5035), 1178–1181 (1991). [CrossRef] [PubMed]
2. S. R. Chinn, E. A. Swanson, and J. G. Fujimoto, “Optical coherence tomography using a frequency-tunable optical source,” Opt. Lett. 22(5), 340–342 (1997). [CrossRef] [PubMed]
3. B. Potsaid, B. Baumann, D. Huang, S. Barry, A. E. Cable, J. S. Schuman, J. S. Duker, and J. G. Fujimoto, “Ultrahigh speed 1050nm swept source/Fourier domain OCT retinal and anterior segment imaging at 100,000 to 400,000 axial scans per second,” Opt. Express 18(19), 20029–20048 (2010). [CrossRef] [PubMed]
4. R. Huber, M. Wojtkowski, K. Taira, J. Fujimoto, and K. Hsu, “Amplified, frequency swept lasers for frequency domain reflectometry and OCT imaging: design and scaling principles,” Opt. Express 13(9), 3513–3528 (2005). [CrossRef] [PubMed]
5. W. Y. Oh, S. H. Yun, G. J. Tearney, and B. E. Bouma, “115 kHz tuning repetition rate ultrahigh-speed wavelength-swept semiconductor laser,” Opt. Lett. 30(23), 3159–3161 (2005). [CrossRef] [PubMed]
6. W.-Y. Oh, B. J. Vakoc, M. Shishkov, G. J. Tearney, and B. E. Bouma, “>400 kHz repetition rate wavelength-swept laser and application to high-speed optical frequency domain imaging,” Opt. Lett. 35(17), 2919–2921 (2010). [CrossRef] [PubMed]
7. M. Kuznetsov, W. Atia, B. Johnson, and D. Flanders, “Compact ultrafast reflective Fabry-Perot tunable lasers for OCT imaging applications,” Proc. SPIE 7554, 75541F (2010). [CrossRef]
8. V. Jayaraman, J. Jiang, H. Li, P. J. S. Heim, G.D. Cole, B. Potsaid, J.G. Fujimoto, and A. Cable, “OCT Imaging up to 760 kHz Axial Scan Rate Using Single-Mode 1310nm MEMS-Tunable VCSELs with >100nm Tuning Range,” presented at CLEO 2011 PDPB2, Baltimore, USA, 1–6 May 2011.
9. M. P. Minneman, J. Ensher, M. Crawford, and D. Derickson, “All-semiconductor high-speed akinetic swept-source for OCT,” Proc. SPIE 8311, 831116, 831116-10 (2011). [CrossRef]
10. R. Huber, M. Wojtkowski, and J. G. Fujimoto, “Fourier domain mode locking (FDML): A new laser operating regime and applications for optical coherence tomography,” Opt. Express 14(8), 3225–3237 (2006). [CrossRef] [PubMed]
11. M. Y. Jeon, J. Zhang, Q. Wang, and Z. Chen, “High-speed and wide bandwidth Fourier domain mode-locked wavelength swept laser with multiple SOAs,” Opt. Express 16(4), 2547–2554 (2008). [CrossRef] [PubMed]
12. S. Marschall, T. Klein, W. Wieser, B. R. Biedermann, K. Hsu, K. P. Hansen, B. Sumpf, K. H. Hasler, G. Erbert, O. B. Jensen, C. Pedersen, R. Huber, and P. E. Andersen, “Fourier domain mode-locked swept source at 1050 nm based on a tapered amplifier,” Opt. Express 18(15), 15820–15831 (2010). [CrossRef] [PubMed]
13. W. Wieser, B. R. Biedermann, T. Klein, C. M. Eigenwillig, and R. Huber, “Multi-Megahertz OCT: High quality 3D imaging at 20 million A-scans and 4.5 GVoxels per second,” Opt. Express 18(14), 14685–14704 (2010). [CrossRef] [PubMed]
14. T. Klein, W. Wieser, C. M. Eigenwillig, B. R. Biedermann, and R. Huber, “Megahertz OCT for ultrawide-field retinal imaging with a 1050nm Fourier domain mode-locked laser,” Opt. Express 19(4), 3044–3062 (2011). [CrossRef]
15. S. Yamashita and M. Asano, “Wide and fast wavelength-tunable mode-locked fiber laser based on dispersion tuning,” Opt. Express 14(20), 9399–9306 (2006). [CrossRef] [PubMed]
16. Y. Nakazaki and S. Yamashita, “Fast and wide tuning range wavelength-swept fiber laser based on dispersion tuning and its application to dynamic FBG sensing,” Opt. Express 17(10), 8310–8318 (2009). [CrossRef] [PubMed]
17. S. Yamashita, Y. Nakazaki, R. Konishi, and O. Kusakari, “Wide and fast wavelength-swept fiber laser based on dispersion tuning for dynamic sensing (invited),” Journal of Sensors 2009, A Special Issue on Fiber and Integrated Waveguide-Based Optical Sensors, 572835 (2009).
18. Y. Takubo and S. Yamashita, “In vivo OCT imaging using wavelength-swept fiber laser based on dispersion tuning,” IEEE Photon. Technol. Lett. 24(12), 979–981 (2012). [CrossRef]
19. B. Burgoyne and A. Villeneuve, “Programmable lasers: design and applications,” Proc. SPIE 7580, 758002, 758002-15 (2010). [CrossRef]
20. G. Lamouche, S. Vergnole, Y. Kim, B. Burgoyne, and A. Villeneuve, “Tailoring wavelength sweep for SS-OCT with a programmable picosecond laser,” Proc. SPIE 7889, 78891L, 78891L-6 (2011). [CrossRef]
21. Y. Kim, B. Burgoyne, N. Godbout, A. Villeneuve, G. Lamouche, and S. Vergnole, “Picosecond programmable laser sweeping over 50 mega-wavelengths per second,” Proc. SPIE 7914, 79140Y, 79140Y-8 (2011). [CrossRef]
22. S. Yamashita and Y. Takubo, “Fast wavelength-swept dispersion-tuned fiber laser over 500kHz using a wideband chirped fiber Bragg grating,” Proc. SPIE 7753, 77537W, 77537W-4 (2011). [CrossRef]
23. H. Don Lee, J. H. Lee, M. Y. Jeong, and C. S. Kim, “Characterization of wavelength-swept active mode locking fiber laser based on reflective semiconductor optical amplifier,” Opt. Express 19(15), 14586–14593 (2011). [CrossRef] [PubMed]
24. F. Lexer, C. K. Hitzenberger, A. F. Fercher, and M. Kulhavy, “Wavelength-tuning interferometry of intraocular distances,” Appl. Opt. 36(25), 6548–6553 (1997). [CrossRef] [PubMed]
25. C. Chong, T. Suzuki, A. Morosawa, and T. Sakai, “Spectral narrowing effect by quasi-phase continuous tuning in high-speed wavelength-swept light source,” Opt. Express 16(25), 21105–21118 (2008). [CrossRef] [PubMed]
26. Santec Corporation, “Inner Vision” OCT system, http://www.santec.com/en/products/oct/octsystem.
27. A. Takada, M. Fujino, and S. Nagano, “Dispersion dependence of linewidth in actively mode-locked ring lasers,” Opt. Express 20(4), 4753–4762 (2012). [CrossRef] [PubMed]
