We present a novel circumferential-scan endoscopic optical coherence tomography (OCT) probe by using a circular array of six electrothermal microelectromechanical (MEMS) mirrors and six C-lenses. The MEMS mirrors have a 0.5 mm × 0.5 mm mirror plate and a chip size of 1.5 mm × 1.3 mm. Each MEMS mirror can scan up to 45° at a voltage of less than 12 V. Six of those mirrors have been successfully packaged to a probe head; full circumferential scans have been demonstrated. Furthermore, each scan unit is composed of a MEMS mirror and a C-lens and the six scan units can be designed with different focal lengths to adapt for lesions with uneven surfaces. Configured with a swept source OCT system, this MEMS array-based circumferential scanning probe has been applied to image a swine’s small intestine wrapped on a 20 mm-diameter glass tube. The OCT imaging result shows that this new MEMS endoscopic OCT has promising applications in large tubular organs.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
12 April 2018: A typographical correction was made to the author affiliations.
Endoscopic optical coherence tomography (OCT) has been drawing a great deal of attention in the optical imaging research community due to its enormous promise in noninvasive early cancer diagnosis and real-time imaging guided tumor removal . One of the critical challenges for endoscopic OCT is the realization of optical beam scanning in an ultra-small form factor . The optical scan modes mainly include forward scanning, side scanning and circumferential scanning . Optical fibers driven by piezoelectric tubes or microelectromechanical (MEMS) microstages can be used to realize forward scanning [1–3]. Side-scanning endoscopic OCT probes are usually attained with two-axis MEMS mirrors to achieve two-dimensional (2D) scans and three-dimensional (3D) OCT images [4–9].
Circumferential scanning was first developed for intravascular OCT , and now it is widely used in clinic . In such a circumferential scanning mechanism, a fiber-GRIN lens-prism module is spun through a guide wire driven by an external motor [12–14]. Meanwhile, circumferential scanning for endoscopic OCT was also demonstrated by using a micromotor to rotate a prism at the distal end of the probe [15–17]. Recently, the spinning guide-wire type of circumferential scanning probes has been utilized in endoscopic OCT for imaging hollow organs with large lumina, such as esophagus and colon [18,19]. This technology is very efficient to acquire circumferential images, but it has difficulty to image a localized specific region of interest.
Raster scanning provided by MEMS mirrors is a good candidate to solve this problem, but almost all the MEMS mirror based endoscopic OCT probes reported in the literature had small working distance and small field of view (FOV) as small-diameter GRIN lenses were typically employed [4–9,17]. Thus it is not practical to apply these MEMS probes for imaging large lumina. On the other hand, C-lenses are also rod lenses that match optical fibers as well as GRIN lenses but can have longer working distance; they have been widely used in optical communications [20–22].
Therefore, in this work, C-lenses can be utilized to extend the working distance. To increase the FOV, one method is to increase the scan angle , but the ruggedness of the MEMS mirror has to be significantly compromised. In this work, we propose to stitch six MEMS mirrors into a circular array to cover the entire full 360° view. In the following, we will first introduce the conceptual design of this novel circumferential-scan MEMS probe, followed by the C-lens design and the MEMS mirror. Then we will describe the final design and assembly process of the circumferential-scan MEMS probe. After that, we will present the probe characterization and OCT imaging results.
In this study, we propose to use a C-lens and a MEMS mirror to compose a scan unit, and six scan units are circular arrayed to be a circumferential scanning probe. The MEMS mirror can scan 45° at 5V offset voltage with ± 7V drive voltage. The focus length of the C-lens is designed at 12 mm in order to image large tubular organs.
2. Materials and methods
2.1 Conceptual design of the circularly arrayed MEMS OCT probe
A schematic of the probe design is illustrated in Fig. 1(a), where six scan units are circularly arrayed and place around a central pole. Each scan unit is composed of an optical fiber, a C-lens collimator, and a MEMS mirror. The MEMS mirrors are placed on a 45° sloped hexagonal prism while the C-lens collimators are attached to the sidewall of the central pole. For each scan unit, the light is coupled through a fiber, collimated by the C-lens, and then focused on the MEMS mirror along the longitudinal direction of the probe. After bouncing off the 45°-tilted MEMS mirror, the light is incident on the lumen surface perpendicularly. The transverse plane scanned by the six optical beams is illustrated in Fig. 1(b), where O1 and O2 are the centers of the probe and one of the MEMS mirrors, respectively, θ is the arc angle corresponding to one sixth of the perimeter of the lumen surface (thus θ = 60°), and is the minimum optical scan angle of each MEMS mirror needed for all six mirrors to cover the full perimeter. Based on simple geometry, , can be expressed asFig. 1(c).
2.2 C-lens design
As shown in Fig. 2(A), a C-lens and a single-mode fiber are packaged inside a glass tube with a diameter of 1.4 mm Note that both the fiber end and the C-lens front end are cut with an 8° angle to minimize back reflection. The air gap between the fiber end and the C-lens front end is defined as b. The output beam from the C-lens rear end is approximately a Gaussian beam. The spot size, ω0, focal point, Z0, and Rayleigh range, ZR, of the output beam can be calculated using ABCD matrix if the diameter, d, length, L, radius of curvature, r, and refractive index, n, of the C-lens and b are given. And the ω0, Z0, and ZR mainly determined by b. Through multiple iterations, the C-lens’ parameters are selected as d = 1 mm, L = 3.2 mm, r = 1.8 mm, and n = 1.74. Then the calculated ω0 and Z0 / ZR over b are shown in Figs. 2(B) and 2(C), respectively. So ω0, Z0 and ZR can be adjusted by changing b.
As shown in Fig. 2(B), b should be longer than 1.0 mm to keep the spot size less than 50 μm. According to Fig. 2(C), b should less than 1.37 mm to ensure the focal length greater than 10 mm. Thus, in this work, b is chosen as 1.2 mm; the corresponding focal length and beam waist diameter are 12.1 mm and 40.3 μm, respectively. The focal length can be changed from 5.0 mm to 14.2 mm while keeping the beam waist less than 50 μm, according to Fig. 2 (B) and (C).
As shown in Fig. 2, the focal length can be changed by adjusting b, the distance between the fiber and the C-lens. Therefore, as shown in Fig. 1(a), there are two scan modes of the circular arrayed probe. The first one is that the six scan units are designed with the same focal length, and then the scanned imaging area will be a continuous uniform circumferential-scan. The second mode is that the six scan units are designed with different focal lengths to match an irregular tissue surface. Figure 1(c) illustrates the multiple scan regions with different imaging depths.
2.3 MEMS mirrors
As shown in Fig. 1(b) and Eq. (1), the smaller the distance between O1 and O2, i.e., r1, is, the smaller the scan angle is required. Smaller r1 means smaller MEMS devices, but small chips typically will have smaller mirror plates, which leads to larger divergence of the light beams and consequentially poorer lateral OCT resolution. In this work, a specially-designed MEMS mirror based on electrothermal bimorph actuation is employed. As shown in Fig. 3(a), four bimorph actuators are symmetrically located at the four sides of a central mirror plate. Each bimorph actuator consists of three pairs of double S-shaped bimorph (Al/SiO2) beams . A thin layer of Ti/TiN is embedded along the bimorph actuator as the heater to drive bimorphs. This MEMS mirror design has high fill factor, i.e., it can keep relatively large mirror plate even at small footprint. An SEM picture of a fabricated MEMS mirror is shown in Fig. 3(a), where the footprint is as small as 1.5 mm × 1.3mm but the mirror plate is still maintained at a moderate size of 0.5 mm × 0.5mm. The device fabrication process is similar to the one reported in .
The measured angular response of the MEMS mirror under a drive voltage is plotted in Fig. 3(b), which is linear between 1.0 V and 7.0 V. A 13° mechanical scan angle is achieved at 7.0 V voltages, resulting in ± 26° optical scan angle or a 52° FOV. Note that this is quasi-static response. At 7 V, the current is about 10 mA. So the heating effect is very small.
The probe is packaged with a glass tube. Image distortion effects of the tubing such as astigmatism were observed and analyzed and signal processing algorithms of correcting those distortions were also reported previously [27–29]. In this work, the beam diameters at different positions out of the probe were measured using a beam analyzer (Thorlabs, BP209-IR); the result is shown in Fig. 4(a), where XOZ is the horizontal plane transverse to the probe, and YOZ is the vertical plane along the probe’s longitudinal direction. It clearly shows that the focal points in the two orthogonal directions were not coincident. The beam waist diameter in XOZ and YOZ of the probe is about 53.1 μm and 52.8 μm, respectively. This can be further seen from the beam profile shown in Fig. 4(b), which was measured at the distance of 12 mm, and the beam width in XOZ is greater than that in YOZ.
3. Manufacture of the circularly arrayed MEMS probe
The final design of the circumferential-scan MEMS probe is shown in Fig. 5(a), where the probe holder consists of a hexagonal prism on the base, a hollow hexagonal pole in the center, and a cylinder wrapping the center pole. The hollow hexagonal pole is used to place the flexible printed circuit board (FPCB). Furthermore, the hexagonal prism’s six sides are all sloped at a 45° angle for placing six MEMS chips while the cylinder has six holes for holding six C-lens collimators. The entire probe holder is 3D printed using stereo lithography apparatus (SLA) technique. The overall dimensions of the probe holder are 25 mm long and 10.2 mm in diameter. The bottom of the hexagonal prism has six grooves to place FPCB to reduce the size.
Six MEMS chips are first wire bonded on six FPCBs respectively, and then fixed on the 45° slopes of the probe holder, as shown in Figs. 6(a) and 6(b). Optical UV glue is used to fix the C-lens collimator to the probe holder when a red light shows good alignment between the C-lens and MEMS chip. Figure 6(c) is a zoomed-in photo showing the C-lens collimators and MEMS chips. All six FPCBs are threaded through the hole of the central hollow pole, as shown in Fig. 6(d), solving the problem of image shadowing by electrical wires in conventional distally-driven circumferential scan probes.
Figure 7(a) shows the assembled probe that has a diameter of 10.2 mm. The length of the probe holder is 25 mm while the fiber pigtail of every C-lens collimator is about 1.2 m long. Figure 7(b) shows the probe packaged into a glass tube with an outer diameter of 12.0 mm. The distance from the outer surface of the glass tube to the focal point of the light beam is about 6~7 mm, as shown in Fig. 7(c).
4. OCT system image experiments and results
The axial resolution of OCT system is determined by the swept source employed in the OCT system, which is a Santec HSL-20-100-B that has a central wavelength of 1310 nm with the full width half maximum (FWHM) of 91 nm, and the swept speed is 100 kHz. The corresponding theoretical axial resolution is 9 μm in air. The actual measured resolution is slightly lower than this theoretical value. In this study, only 1D scan was performed and the drive voltage signal was a 5 Hz triangular waveform. The actual optical scan angle range at 5 Hz was about 45°, slightly smaller than the static 52°. This is caused by the relatively large thermal response time (~20 ms) of the MEMS mirror [4,24]. However, MEMS mirrors with more than 60° optical scan angle are achievable with modifying the actuator designs or simply operating at the resonant frequency . A full B-scan of each scan unit takes 200 ms. Each A-scan takes 10 μs. So there are totally 20,000 A-scans for each B-scan. Repetitive B-scans are typically employed to obtain higher quality images.
A two-layer coverslips sample was made to test the characteristics of the scan units. As shown in Fig. 8(a), a bare fiber (diameter = 125 μm) is clamped within two coverslips (thickness = 165 μm), producing an air wedge with a 125 μm gap at the end. In this way, the imaging position and the air gap can be precisely controlled. The top and bottom surface of each coverslip are imaged, so there are four reflecting surfaces. Multiple A-scans were taken at a single point near the fiber; the corresponding reconstructed OCT image is shown in Fig. 8(b1). One A-scan signal is plotted in Fig. 8(b2), where the 22 pixels between 1st and 2nd peaks or between the 3rd and 4th peaks represent 165 μm in the glass (whose refractive index is 1.51) while the 11 pixels between 2nd and 3rd peak represent 125 μm in air. So the sensitivity of pixel in the glass and air are 7.5 μm per pixel and 11.3 μm per pixel, respectively, and their ratio (1.5) is the refractive index of glass. A complete B-scan was performed at the location where the air gap is about 12 μm, which is near the resolution limit of the OCT system. The corresponding OCT image of the full B-scan is shown in Fig. 8(c1), where the flat surfaces appear curved but the 12 μm air gap is clearly delineated. The image distortion results from the radial scan of the MEMS mirror, forming spherical, fan-shaped and keystone distortions that can be corrected by image-revers methods . Figure 8 (c2) shows an A-scan line signal, where the air gap is presented as a single pixel with low intensity.
Figure 9 shows the architecture of the OCT system with the arrayed MEMS probe, in which a 1 × 6 MEMS optical switch (SW 1 × 6-9N, HTX Photonics CO., LTD, Wuhan, China) is used to deliver the light to the six scan units sequentially and the switch time from one unit to another is less than 0.5 ms. As the B-scan of each scan units takes 200 ms, the total time for taking a full circumferential image is 1.203 s. As shown in Fig. 10 (b), the six optical channels are combined to six scan units of the probe. Each scan unit acquires data independently and the data from all six scan units can be used to reconstruct integrated OCT image of tubular samples.
A sample was prepared by wrapping a swine’s small intestine around a glass tube with an outer diameter of 23 mm, and the OCT system with the arrayed MEMS probe was used to image this sample, as shown in Figs. 10(a) and 10(b). A reconstructed OCT image is shown in Fig. 10(c), where each scan unit of this probe scanned 45° in this experiment. So the OCT image consists of six radial sections of images and the scan center of each image section is at the center of the corresponding MEMS mirror. However, the OCT image is not a single continuous annulus but separated by six shadow gaps. A minimal 60° scan angle from each scan unit is needed to form a continuous OCT image. This can be achieved by developing new MEMS mirrors with larger scan angles. MEMS mirrors based on the electro-thermal bimorph actuator design employed in this MEMS mirror have shown over 60° scan FOV [6,24].
5. Discussion and conclusion
One interesting feature of the proposed probe is that it can generate shadow-free images as the electrical wires are folded and threaded through the central hole in the probe. However, as we can see from Fig. 10, there are still six shadow areas in the OCT image. This is not caused by the probe design but rather by the limited scan angle (45° instead of needed 60° FOV) of the MEMS mirror. There are two methods to solve this problem. Firstly, we can design and fabricate a new MEMS mirror with lager scan angle. Secondly, we can use eight or more scan units in the probe to cover the full 360° view. Another interesting feature of this probe is that each scan unit is independent and the six scan units can be designed with totally different scan radii to match various surface profiles of lesions or combine two or three scan units together to balance the scan range and variety.
Furthermore, this probe has a potential advantage of acquiring 3-D cylindrical images without pushing or pulling the probe as the MEMS can scan the light beam along the probe too. The stationary probe results in a more stable and higher quality image. As each scan unit is independent, the probe is more controllable. For example, we can change the circumferential scan mode to a more localized scan mode by simply using one scan unit during test. This feature will be investigated in the near future.
In summary, in this work, we have successfully demonstrated a 12 mm-diameter packaged endoscopic OCT probe based on six 2-axis electro-thermal MEMS mirrors and six C-lenses. Every scan unit is composed of a MEMS mirror and a C-lens. The six scan units are circularly arrayed to realize circumferential scan. The MEMS mirrors can scan 45°at 12 V. The diameter of the circumferential imaging range reaches as much as 20 mm. The focal length of each scan unit can be variously designed to fit tubular organs that do not have perfect circular cross sections. Imaging results of the novel arrayed MEMS probe based on a swept source OCT system show that this full circumferential scan OCT is a promising technology for endoscopic in vivo diagnosis for large tubular organs. In the further study, more systematic investigation of the imaging performance of this new probe, including 3D imaging, astigmatism, uniformity among the six scan units, and imaging on tubular surfaces with varied curvature, will be performed. Also we plan to reduce the packaged probe diameter to 7-8 mm by optimizing the structure of the probe holder and utilizing organic glass tube with less wall thickness.
National Natural Science Foundation of China (No. 61575107, No. 61528401); Ministry of Science and Technology of China's Key Project of Research and Development Plan (2017YFC0108301); US National Science Foundation (1512531).
The authors declare that there are no conflicts of interest related to this article.
References and links
1. M. J. Gora, M. J. Suter, G. J. Tearney, and X. Li, “Endoscopic optical coherence tomography: technologies and clinical applications [Invited],” Biomed. Opt. Express 8(5), 2405–2444 (2017). [CrossRef] [PubMed]
3. S. A. Boppart, B. E. Bouma, C. Pitris, G. J. Tearney, J. G. Fujimoto, and M. E. Brezinski, “Forward-imaging instruments for optical coherence tomography,” Opt. Lett. 22(21), 1618–1620 (1997). [CrossRef] [PubMed]
4. L. Liu, L. Wu, J. Sun, E. Lin, and H. Xie, “Miniature endoscopic optical coherence tomography probe employing a two-axis microelectromechanical scanning mirror with through-silicon vias,” J. Biomed. Opt. 16(2), 026006 (2011). [CrossRef] [PubMed]
5. W. Jung, D. T. McCormick, J. Zhang, L. Wang, N. C. Tien, and Z. Chen, “Three-dimensional endoscopic OCT by use of a two-axis MEMS scanning mirror,” Appl. Phys. Lett. 88, 163901 (2006). [CrossRef]
6. J. Sun, S. Guo, L. Wu, L. Liu, S. W. Choe, B. S. Sorg, and H. Xie, “3D in vivo optical coherence tomography based on a low-voltage, large-scan-range 2D MEMS mirror,” Opt. Express 18(12), 12065–12075 (2010). [CrossRef] [PubMed]
7. K. H. Kim, B. H. Park, G. N. Maguluri, T. W. Lee, F. J. Rogomentich, M. G. Bancu, B. E. Bouma, J. F. de Boer, and J. J. Bernstein, “Two-axis magnetically-driven MEMS scanning catheter for endoscopic high-speed optical coherence tomography,” Opt. Express 15(26), 18130–18140 (2007). [CrossRef] [PubMed]
8. D. Wang, L. Fu, X. Wang, Z. Gong, S. Samuelson, C. Duan, H. Jia, J. S. Ma, and H. Xie, “Endoscopic swept-source optical coherence tomography based on a two-axis microelectromechanical system mirror,” J. Biomed. Opt. 18(8), 086005 (2013). [CrossRef] [PubMed]
10. G. J. Tearney, M. E. Brezinski, B. E. Bouma, S. A. Boppart, C. Pitris, J. F. Southern, and J. G. Fujimoto, “In vivo endoscopic optical biopsy with optical coherence tomography,” Science 276(5321), 2037–2039 (1997). [CrossRef] [PubMed]
11. Nine Point Companyhttp://www.ninepointmedical.com/nvisionvle-imaging-system
13. B. E. Bouma and G. J. Tearney, “Power-efficient nonreciprocal interferometer and linear-scanning fiber-optic catheter for optical coherence tomography,” Opt. Lett. 24(8), 531–533 (1999). [CrossRef] [PubMed]
14. V. X. D. Yang, Y. X. Mao, N. Munce, B. Standish, W. Kucharczyk, N. E. Marcon, B. C. Wilson, and I. A. Vitkin, “Interstitial Doppler optical coherence tomography,” Opt. Lett. 30(14), 1791–1793 (2005). [CrossRef] [PubMed]
15. P. R. Herz, Y. Chen, A. D. Aguirre, K. Schneider, P. Hsiung, J. G. Fujimoto, K. Madden, J. Schmitt, J. Goodnow, C. Petersen, and C. Petersen, “Micromotor endoscope catheter for in vivo, ultrahigh-resolution optical coherence tomography,” Opt. Lett. 29(19), 2261–2263 (2004). [CrossRef] [PubMed]
16. J. Li, M. de Groot, F. Helderman, J. Mo, J. M. A. Daniels, K. Grünberg, T. G. Sutedja, and J. F. de Boer, “High speed miniature motorized endoscopic probe for optical frequency domain imaging,” Opt. Express 20(22), 24132–24138 (2012). [CrossRef] [PubMed]
17. N. Zhang, T.-H. Tsai, O. O. Ahsen, K. Liang, H.-C. Lee, P. Xue, X. Li, and J. G. Fujimoto, “Compact piezoelectric transducer fiber scanning probe for optical coherence tomography,” Opt. Lett. 39(2), 186–188 (2014). [CrossRef] [PubMed]
18. T.-H. Tsai, H.-C. Lee, O. O. Ahsen, K. Liang, M. G. Giacomelli, B. M. Potsaid, Y. K. Tao, V. Jayaraman, M. Figueiredo, Q. Huang, A. E. Cable, J. Fujimoto, and H. Mashimo, “Ultrahigh speed endoscopic optical coherence tomography for gastroenterology,” Biomed. Opt. Express 5(12), 4387–4404 (2014). [CrossRef] [PubMed]
20. W. Jing, D. Jia, F. Tang, H. Zhang, Y. Zhang, G. Zhou, J. Yu, F. Kong, and K. Liu, “Design and implementation of a broadband optical rotary joint using C-lenses,” Opt. Express 12(17), 4088–4093 (2004). [CrossRef] [PubMed]
21. CASIX CO, LTD, “Collimators: C-lens”,http://www.casix.com/products/fiber-optics-subs/collimators.shtml
22. S. Wang, Y. Ruan, D. Yin, and Z. Yang, “The calculation and analyzing of the RL of C-lens collimator,” Optoelectronic Technology and Information 16(1), 24–28 (2003).
23. L. Wu and H. Xie, “Electrothermal micromirror with dual-reflective surfaces for circumferential scanning endoscopic imaging. J. of Micro/Nanolithography,” MEMS, and MOEMS 8(1), 013030 (2009). [CrossRef]
24. S. R. Samuelson, L. Wu, J. J. Sun, S. W. Choe, B. S. Sorg, and H. K. Xie, “A 2.8-mm Imaging Probe Based On a High-Fill-Factor MEMS Mirror and Wire-Bonding-Free Packaging for Endoscopic Optical Coherence Tomography,” J. Microelectromech. Syst. 21(6), 1291–1302 (2012). [CrossRef]
25. D. Wang, P. Liang, S. Samuelson, H. Jia, J. Ma, and H. Xie, “Correction of image distortions in endoscopic optical coherence tomography based on two-axis scanning MEMS mirrors,” Biomed. Opt. Express 4(10), 2066–2077 (2013). [CrossRef] [PubMed]
26. S. R. Samuelson and H. Xie, “A Large Piston Displacement MEMS Mirror With Electrothermal Ladder Actuator Arrays for Ultra-Low Tilt Applications,” J. Microelectromech. Syst. 23(1), 39–49 (2014). [CrossRef]
27. T. Wang, A. F. W. van der Steen, and G. van Soest, “Numerical analysis of astigmatism correction in gradient refractive index lens based optical coherence tomography catheters,” Appl. Opt. 51(21), 5244–5252 (2012). [CrossRef] [PubMed]
29. C. Duan, J. Sun, S. Samuelson, and H. Xie, “Probe alignment and design issues of microelectromechanical system based optical coherence tomography endoscopic imaging,” Appl. Opt. 52(26), 6589–6598 (2013). [CrossRef] [PubMed]