Open Access
Add to BibSonomy
Abstract
Micro–electro–mechanical systems (MEMS)-based optical scanners play a vital role in the development of miniaturized optical imaging modalities. However, there is a longstanding challenge to balance the temporal resolution, field of view (FOV), and systematic fidelity. Here, we propose a double spiral scanning mechanism to enable high-frequency resonant scanning of MEMS scanners without sacrificing imaging quality, and offer a versatile imaging interface for applications in different scenarios. This arrangement, demonstrated by photoacoustic endoscopy, shows that the imaging rate and FOV can be improved by more than 60 and two times, respectively. The proposed method is general to address the limitations of MEMS-based scanning microscopies and can be adapted for various miniaturized imaging modalities, such as endoscopy, intraoperative image-guided surgery, and wearable devices.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. INTRODUCTION
Miniaturized optical imaging modalities, especially laser scanning microscopies, pave the way for progressing the diagnostic endoscopy [1–7], advancing image-guided surgery [8–10], and developing wearable devices [11–16] in various biomedical studies due to the superior features of low cost, flexibility, high spatiotemporal resolution, and multi-contrast. However, one of the major challenges restricting the performance of miniaturized laser scanning microscopies is the scanning device and mechanism. The advancement of the micro–electro–mechanical systems (MEMS) scanner, serving as the primary scanning device inside state-of-the-art miniaturized imaging probes, has achieved as small as a few millimeters in dimension [17,18]. Unfortunately, MEMS scanners are severely affected by the limited tilting angle and small mirror plate, resulting in an insufficient field of view (FOV) and low spatial resolution, respectively. Additionally, the poor stability of MEMS at high driving frequencies constitutes another crucial factor that restricts the imaging quality of miniaturized imaging probes.
To date, most MEMS-based imaging probes employ raster scanning or Lissajous scanning schemes [19,20]. The raster scanning scheme typically uses a triangular or sawtooth scanning pattern to drive the focused beam point by point in the imaging field. However, the imaging rate is limited, as it suffers from scan distortion, non-uniform scanning at a high driving frequency, and a deadtime for flyback. Optimization of driving control [21–24] and utilization of more advanced material processes [25] can both enhance scanning linearity, speed, and stability of MEMS scanners. Nevertheless, it requires a brand-new design of the MEMS scanner and cannot be directly implemented on general MEMS scanners. Modified scanning patterns such as sinusoidal scanning patterns [11–13,15,26,27] and spiral scanning patterns [28,29] can overcome the sharp turning issue and provide smoother and faster scanning, yet at the cost of a reduced linear scan region and suffering from the step response deadtime and natural braking. Recently, continuous scanning schemes [30–33] have been proposed to overcome the non-continuous acquisition. These methods enable the MEMS scanner to work at a higher driving frequency, and thus lead to faster scanning. However, the primary limitation is the inconsistent frequency response during scanning, which requires a point-by-point calibration for image reconstruction. In addition to the aforementioned issues, another critical limitation of these conventional scanning schemes is the deformation of the scan mirror that severely deteriorates image quality when increasing the driving frequency. These factors are inter-connected, as shown in Fig. 1(a), hindering MEMS-based imaging systems in acquiring large FOV, high-quality images at a fast frame rate.
Fig. 1. Mechanism of double spiral resonant scanning (DSRS). (a) Schematic diagram of the scanning mechanism. The light beam is deflected by the MEMS scanner with an incidence angle of $\beta$. Parameters of scanning frequency, deflection angle, and size of the MEMS mirror limit the optimal imaging performance. The total length of the dotted line in the middle of the optical path is the scanning distance (SD). (b) The scanning angle directly determining the FOV is frequency dependent. The scanning angle decreases with the increase of the driving frequency in the non-resonance regime. It dramatically increases when reaching the resonant frequency range (red shaded background). The red circles indicate the FOVs at the corresponding driving frequencies. The horizontal dotted line is the damage threshold. (c) Image distortion is affected by three factors: smoothness of the scanning pattern (I, blue), step response of the MEMS scanner (II, pink), and dynamic deformation of the mirror (III, black). (d) Schematic diagram of double spiral scanning trajectory. (e) Change of acceleration in double spiral scanning. (f) The DSRS technique provides an optimized solution to address the inter-connected constraints of imaging speed, FOV, and fidelity in MEMS-based scanning.
In this work, we introduce a double spiral resonant scanning (DSRS) scheme to realize the goals of large and distortion-free FOV at a high-speed resonant frequency for miniaturized laser scanning microscopies. We theoretically analyzed the inter-connected relationship of MEMS mirror deformation, driving frequency, and FOV, and propose an optimized MEMS-based scanning scheme. Using an electrothermal MEMS scanner, we validated the DSRS and showed that it significantly improved the stable driving frequency from 50 Hz to the resonant peak of 780 Hz, and increased the tilting angle of ${\pm}{4}\;{\rm deg}$ to ${\pm}\;{6}\;{\rm deg}$ without sacrificing image quality. We evaluated the DSRS for fast, large FOV, and deformation-free imaging using miniaturized photoacoustic microscopy (PAM), optical coherence tomography (OCT), and fluorescence confocal microscopy.
2. CONCEPT OF THE DSRS SCHEME
It has been challenging for MEMS-based scanning to achieve fast imaging, large FOV, and high image quality all at once since it requires a combination of three exclusively constrained features at the same time including a high driving frequency, large tilting angle, and fidelity [19,20]. Figure 1(b) shows the frequency response curve of the electrothermal MEMS scanner (OC-MI-T10, Wuxi WiO Technology Co., Ltd.). It is worth noting that almost all MEMS scanners have a specific frequency response curve and a maximum scanning angle that depends on the driving frequency and directly determines the FOV. Before reaching the resonant range, the scanning angle decreases with the increase of the driving frequency. With further increase of the driving frequency to the intrinsic resonant frequency range, we see that the scanning angle increases dramatically, which provides a window to realize large FOV and fast imaging speed at the same time. However, there exists an unstable zone near the resonant frequency range [15] [highlighted by the red shaded region in Fig. 1(b)], within which a slight perturbation of the driving signal can cause an amplified scanning error as well as severe deterioration of image quality.
In addition, it becomes increasingly difficult to obtain high-fidelity images at fast driving frequencies. As illustrated in Fig. 1(c), image quality is mainly affected by three factors, including abrupt changes of the scanning speed (I), step response of MEMS scanners (II), and dynamic deformation of the mirror (III). The utilization of a smooth and continuous scanning pattern can effectively mitigate the influence of the first two factors and enable improvement in imaging speed without sacrificing image quality. However, as the driving frequency increases, the mechanical deflection of the mirror is exacerbated by strong acceleration forces, leading to serious deformation of the mirror surface. The dynamic deformation of the mirror ${\delta _{{\max}}}$, defined as the deviation from linearity, can be depicted using Brosens’s formula [19] (see Supplement 1, Note 1 for more derivation)
where $\theta (t)$ is the maximum mechanical tilting angle in the scan circle at time $t$, $\omega$ is the driving frequency, $\rho$ represents the density of the mirror, $h$ is the mirror thickness, $a$ is half of the mirror length, and $E$ is the modulus of elasticity. In fact, the deformation of the MEMS scanner is subject to several factors, including the driving process, material properties, and mirror thickness. In the case of a specific MEMS scanner, the inherent influencing factors (materials, shape, thickness, etc.) are determined. Nonetheless, by concentrating on optimizing the scanning methods, it is feasible to effectively minimize mirror deformation during scanning. Equation (1) suggests that mirror deformation (${\delta _{{\max}}}$) depends on the scanning angle and acceleration force associated with mechanical deflection. Therefore, a scanning scheme with a continuous, smooth trajectory and minimal acceleration force is preferable to achieve high-speed and stable scanning.Here, we propose a DSRS scheme to achieve the goal of fast imaging speed, maximum scanning angle, and minimized mirror deformation through a smooth, continuous scan with a driving frequency in the resonant range. The trajectory of the DSRS scheme is defined as
The scanning pattern of the DSRS scheme is shown in Fig. 1(d). It consists of outward spiral scanning starting from the center followed by inward spiral scanning ending at the center. Thus, the scanning trajectory is smooth and continuous throughout the entire course of imaging, which can not only enable scanning at the resonant frequency, providing fast scanning speed and large scanning angle, but also reduce average mirror deformation as the scan angle is gradually increasing [Fig. S1(b) in Supplement 1]. In addition, the DSRS scheme requires a smaller driving torque to move the mirror as the acceleration is minimized during each scan cycle (see Supplement 1, Note 2 for details). Therefore, the DSRS scheme can overcome the instability near the resonant frequency and enable high-speed stable scanning [Fig. 1(e) and Fig. S1(c) in Supplement 1].
3. RESULT
A. Performance Evaluation of the DSRS Scheme
We built an adjustable photoacoustic microscope platform to evaluate the performance of different scanning schemes [Fig. 2(a); see Supplement 1, Extended Methods–PAM platform for MEMS evaluation]. Figure 2(b) compares the results obtained with the widely used triangle and sawtooth raster scanning schemes and the proposed DSRS scheme. We see that the former two scanning schemes can provide high-quality results at low driving frequencies [Fig. S2 in Supplement 1], yet suffer from severe distortion when the driving frequency increases to ${\sim}{100}\;{\rm Hz}$ [Fig. 2(b)]. As shown in Fig. S3 in Supplement 1, the images obtained from sinusoidal raster and Lissajous scanning schemes also produce severe deformation in high-speed scanning. In contrast, the DSRS scheme provides high-quality images for all driving frequencies, and the images obtained by inward spiral scanning and outward spiral scanning are identical, as shown in Fig. S4 in Supplement 1. Further, we compared different spiral scanning schemes including single spiral scanning, start–end spiral scanning, and DSRS, as shown in Fig. 2(c). We see that high-quality images are obtained for all three spiral scanning schemes at the driving frequency of 100 Hz, while the single spiral scanning scheme (I) suffers from strong distortion in the middle area when the driving frequency is over 200 Hz, and the start–end spiral scanning scheme (II) bears deteriorated image quality at the driving frequency of 780 Hz (dashed white box). In contrast, the DSRS scheme preserves consistent image quality at all driving frequencies. The driving signals for the MEMS scanner in various spiral scanning modes are depicted in Fig. S5 in Supplement 1. Figure S6 in Supplement 1 further indicates the optimal driving frequencies for different scanning schemes. Table S1 in Supplement 1 summarizes the maximum driving frequencies for different scanning schemes with the acceptable image quality.
Fig. 2. DSRS performance evaluation. (a) Experimental setup of photoacoustic microscopy (PAM) platform for MEMS evaluation. L, lens; PH, pinhole; SMF, single mode fiber; SL, scanning lens; TL, tube lens; Obj, objective lens; UT, ultrasonic transducer. (b) PAM images of the grid plate (top row) and the vasculature inside a mouse intestine (bottom row) obtained with triangle, sawtooth, and double spiral scanning patterns at a driving frequency of 100 Hz. Scale bar: 1 mm. (c) PAM images of a mouse cecum obtained with the single spiral scanning, start–end spiral scanning, and double spiral scanning patterns at driving frequencies of 100 Hz, 200 Hz, and 780 Hz. All three spiral scanning modes can work stably at the driving frequency of 100 Hz, while only the DSRS scheme achieves distortion free acquisition for the driving frequency of 780 Hz. See Visualization 1 for dynamic demonstration of these scanning methods. (d) In vivo mice cecum imaging at driving frequencies of 50 Hz, 100 Hz, 200 Hz, 400 Hz, and 780 Hz using the DSRS scheme.
Figure 2(d) compares the maximum FOV of the DSRS scheme at different driving frequencies with a fixed scanning distance of 10 mm. We see that the FOV decreases before the driving frequency reaches the MEMS resonant range (${\sim}{760}\;{\rm Hz}$), and significantly increases at 780 Hz. Therefore, the DSRS scheme can not only achieve high-speed scanning, but also provide an enlarged FOV. Figure S7 in Supplement 1 presents additional images of a scale board acquired at different driving frequencies.
B. DSRS-Based Photoacoustic Endoscopy for Video-Rate in vivo Imaging
It has been challenging to acquire high-quality microscope images from living animals or humans, as the irresistible motion due to heartbeat or breathing severely distorts the sampling field. We developed a MEMS-based miniaturized photoacoustic endoscope (PAE) to demonstrate that the high-speed scanning ability provided by the DSRS scheme can mitigate this severe motion issue and enable distortion-free, large FOV imaging.
Figure 3(a) shows the schematic of a PAE that was developed based on an electrothermal MEMS scanner. The endoscope was designed to have an incident beam angle of 22.5 deg to minimize scanning aberrations, and a folded optical path design to reduce the size of the imaging probe, which also provides flexibility for variable-view endoscope design (see Supplement 1, Note 3 for details). The developed endoscope has an outer diameter of 8 mm and circular FOV of 3 mm in diameter using a MEMS scanner with a 1 mm diameter mirror plate. In addition, we employed a sparse sampling strategy to further reduce the sampling points for each scanning trail, which not only enables a higher volume rate but also reduces laser irradiation to the subject. To address the issues of reduced signal-to-noise ratio (SNR) and deteriorated imaging resolution, we used a deep-learning algorithm based on the pix2pix framework to recover the image. The reconstruction algorithm was derived from a conditional generative adversarial network (cGAN) [34,35] (see Supplement 1, Note 4 for details). With the DSRS scheme and sparse scanning, we achieved a high temporal resolution of 12 volumes/s over a 3 mm circular FOV at a spatial resolution of 7.45 µm.
Fig. 3. In vivo imaging results of the rat abdominal cavity. (a) Schematic of the PAE probe. (b) Schematic diagram of the moving trajectory in in vivo experiments and the stitching process of acquired images. (c) Stitched PAM result of a rat stomach. The inset in the lower right corner shows a photograph of the stomach. (d) Photoacoustic results obtained with traditional raster scanning method (left, frame rate limited to 0.2 Hz) and the DSRS (right) of the rat stomach within white circle region in (c). See Visualization 8 for a comparison between traditional raster scanning and DSRS. Also, see Visualization 9, which shows the real-time imaging of the mouse cecum and stomach. (e) Relative movement of blood vessels in the inner and outer layers of the rat stomach is observed. White circles indicate the movement of superficial vessels, while blue circles indicate the movement of inner layer vessels. (f) Stitched photoacoustic image of a rat intestine. The inset in the upper right corner shows a photograph of the intestine. Scale bar: 1 mm.
To evaluate the capability of DSRS-PAE for high-speed in vivo imaging, we acquired data from living rats [Fig. 3] and performed experiments to image the blood vessels in human oral cavities [Fig. 4]. For animal imaging, the rats were immobilized on a platform and imaged by sweeping the handheld probe over the region of interest. Then, the acquired images were stitched together to obtain a larger FOV [Fig. 3(b)]. Figure 3(c) shows the stitched photoacoustic image of a living rat stomach. Visualization 2 shows the selected close-up views of the full-size MAP image of the stomach in different locations. We noticed serious motion due to the heart beating and breathing, which was a killing factor to obtain high-quality images with laser scanning microscopies. The high-speed imaging capability of the DSRS-PAE can greatly address this problem, as shown in Fig. 3(c) and Visualization 3, where the images obtained are stable during the acquisition period. It is worth noting that the imaging region may become out of the optical plane when moving the probe, leading to variated SNR and resolution. Figure 3(d) compares the results obtained with traditional raster scanning (top) and the DSRS scheme (bottom). We can see that the former suffers from strong motion artifacts, while the latter is able to completely eliminate such artifacts. With longitudinal observation at one location, we captured the delamination movement of superficial mucosal vessels and internal vessels of the stomach as shown in Fig. 3(e) (white circles, relative movement of the outer vessels; blue circles, relative movement of the inner layer vessels). Visualization 4 presents the movement of the vessels during the entire course of the experiment. Figure 3(f) shows the stitched photoacoustic image of a rat intestine. Visualization 5 shows the entire stitching process. In addition, as shown in Fig. S8 in Supplement 1, the DSRS-PAE allows us to acquire images with a high spatiotemporal resolution for various organs of the rodent digestive system including intestine, cecum, and stomach of a mouse. Figure S9 in Supplement 1 presents the stitched images of a mouse cecum.
Fig. 4. In vivo imaging results of a healthy male cavity. (a) Stitched photoacoustic image of the upper lip. (b) Blood vessels in the tip and middle regions of the tongue. Scale bar: 0.5 mm. (c) Blood vessels in the gum and surrounding mucous membranes. The white dotted line shows the boundary of the gum. Scale bar: 0.5 mm. (d) Stitched photoacoustic images obtained at different regions of the lower lip. Scale bar: 1 mm.
To further evaluate the capability of the proposed technique for clinical use, we acquired photoacoustic images of the oral cavity from healthy adult volunteers. The miniaturized imaging probe can be easily placed to the targeting locations, including tongue, gum, and upper and lower lips [Fig. 4, Visualization 6 and Visualization 7]. Figure 4(a) presents the stitched image of the upper lip. Figure 4(b) shows the vasculatures at different regions of the tongue. Figure 4(c) shows the blood vessels within the gum and surrounding tissues, where a distinct transitional border was observed between the gingiva and the mucosa. From the lower lip result, as shown in Fig. 4(d), we observed different patterns of vasculatures, progressing from fine and pointy in the outer lip to thicker and more stump-like in the inner lip. The high-resolution, large FOV imaging ability of the DSRS-PAE can potentially enable the detection and diagnosis of a range of disorders of the oral mucosa, such as ulceration and inflammation [36,37]. Particularly, the fast imaging ability can not only enable stitched images for a larger region that can be canvassed for early detection of suspicious lesions and non-invasive follow-up examination, but also acquire images from organs that are difficult to stabilize, such as the tongue.
4. DISCUSSION
In this study, we proposed a DSRS scheme to address the general challenges of slow imaging rate, small FOV, and image distortion when using a MEMS scanner. The DSRS scheme features continuous, smooth scanning at the resonant frequency, which enables MEMS scanners to realize the goal of a distortion-free, large FOV and high frame rate. The inherent qualities of DSRS, including gradual alteration of scanning angle and acceleration and uniform driving frequency of both axes, impart to the scanning scheme extensive versatility and enhanced stability.
Our results based on PAM indicate that the imaging speed can be increased by approximately 60 times, and the imaging FOV is twice enlarged without sacrificing image quality. We notice that the proposed concept holds significant generalizability and can be adapted for diverse optical imaging modalities. In addition to PAM, we further applied the method in OCT and laser scanning fluorescence microscopy (LSFM) to highlight its general impact in the field. As shown in Fig. S10 in Supplement 1, we utilized the similar design in PAM and assembled two front-view endoscopic imaging probes for OCT and LSFM using the same electrothermal MEMS scanner. Without using sparse sampling, both probes have the same imaging rate of 3 volumes/s and a FOV of 3 mm. As shown in Fig. S11 in Supplement 1, we imaged the finger using the OCT probe and microspheres using the LSFM probe.
We demonstrated that the DSRS scheme provides significant improvement in performances for endoscope probes. In addition to endoscopy, the features of the developed methodology, including miniaturization, light weight, high spatial resolution, fast imaging speed, and large FOV, can be employed to address issues in the development of wearable devices and intraoperative image-guided tools.
Besides biomedical optical imaging modalities, the DSRS scheme can also be used in other MEMS-scanner-based systems, such as light detection and ranging (LiDAR) and pico-projectors. Additionally, as a scanning scheme, the DSRS can optimize any 2D optical scanners such as various MEMS scanners including electrostatic, electrothermal, and electromagnetic MEMS scanners, and galvanometer-based scanners. We believe that the proposed DSRS technique will generate wide impact for the development of laser-scanning-based imaging technologies.
Funding
National Natural Science Foundation of China (61528401, 61775028, 62022037, 62105140, 81571722); Guangdong Science and Technology Department (2019ZT08Y191, SZBL2020090501013); Guangdong Provincial Department of Education (2021ZDZX1064); Guangdong Provincial Key Laboratory of Advanced Biomaterials (2022B1212010003); Shenzhen Science and Technology Program (JCYJ20200109141222892, JCYJ20220818101403008, KQTD20190929172743294, RCBS202110706092213005); Start-up grant from Southern University of Science and Technology.
Acknowledgment
We thank the volunteers who took part in this study. We acknowledge support from Wuxi WiO Technology for technological assistance on the MEMS scanner. L.X. conceived this concept and supervised the overall project design and execution. L.L. and X.L. designed the experiments. L.L. and Wz.Q. developed the control program. X.L., H.G. and T.J. designed and assembled the PAM probe as well as prepared the animal model for in vivo experiments. Human in vivo imaging was performed by L.L., X.L., and H.G. L.L. developed the neural network framework. W.Q. performed the experiments of OCT and confocal laser scanning fluorescence microscopy. L.L., X.L., J.T., and L.X. drafted the paper. All authors provided critical feedback to the paper.
Disclosures
The authors declare no conflicts of interest.
Data availability
The main data supporting the results in this study are available within the paper and Supplement 1. The raw datasets are too large to be publicly shared, yet they are available for research purposes upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
REFERENCES
1. D. K. Pleskow, M. S. Sawhney, P. K. Upputuri, T. M. Berzin, M. F. Coughlan, U. Khan, M. Glyavina, X. Zhang, L. Chen, C. J. Sheil, J. M. Cohen, E. Vitkin, Y. N. Zakharov, I. Itzkan, L. Zhang, L. Qiu, and L. T. Perelman, “In vivo detection of bile duct pre-cancer with endoscopic light scattering spectroscopy,” Nat. Commun. 14, 109 (2023). [CrossRef]
2. H. Guan, D. Li, H. C. Park, A. Li, Y. Yue, Y. A. Gau, M. J. Li, D. E. Bergles, H. Lu, and X. Li, “Deep-learning two-photon fiberscopy for video-rate brain imaging in freely-behaving mice,” Nat. Commun. 13, 1534 (2022). [CrossRef]
3. J. M. Yang, C. Favazza, R. Chen, J. Yao, X. Cai, K. Maslov, Q. Zhou, K. K. Shung, and L. V. Wang, “Simultaneous functional photoacoustic and ultrasonic endoscopy of internal organs in vivo,” Nat. Med. 18, 1297–1302 (2012). [CrossRef]
4. H. Guo, C. Song, H. Xie, and L. Xi, “Photoacoustic endomicroscopy based on a MEMS scanning mirror,” Opt. Lett. 42, 4615–4618 (2017). [CrossRef]
5. Y. Liang, W. Fu, Q. Li, X. Chen, H. Sun, L. Wang, L. Jin, W. Huang, and B. O. Guan, “Optical-resolution functional gastrointestinal photoacoustic endoscopy based on optical heterodyne detection of ultrasound,” Nat. Commun. 13, 7604 (2022). [CrossRef]
6. W. Yuan, R. Brown, W. Mitzner, L. Yarmus, and X. Li, “Super-achromatic monolithic microprobe for ultrahigh-resolution endoscopic optical coherence tomography at 800 nm,” Nat. Commun. 8, 1531 (2017). [CrossRef]
7. A. Lombardini, V. Mytskaniuk, S. Sivankutty, E. R. Andresen, X. Chen, J. Wenger, M. Fabert, N. Joly, F. Louradour, A. Kudlinski, and H. Rigneault, “High-resolution multimodal flexible coherent Raman endoscope,” Light Sci. Appl. 7, 10 (2018). [CrossRef]
8. F. LaRocca, D. Nankivil, T. DuBose, C. A. Toth, S. Farsiu, and J. A. Izatt, “In vivo cellular-resolution retinal imaging in infants and children using an ultracompact handheld probe,” Nat. Photonics 10, 580–584 (2016). [CrossRef]
9. H. Ramakonar, B. C. Quirk, R. W. Kirk, J. Li, A. Jacques, C. R. P. Lind, and R. A. McLaughlin, “Intraoperative detection of blood vessels with an imaging needle during neurosurgery in humans,” Sci. Adv 4, eaav4992 (2018). [CrossRef]
10. W. Yuan, D. Chen, R. Sarabia-Estrada, H. Guerrero-Cazares, D. Li, A. Quinones-Hinojosa, and X. Li, “Theranostic OCT microneedle for fast ultrahigh-resolution deep-brain imaging and efficient laser ablation in vivo,” Sci. Adv 6, eaaz9664 (2020). [CrossRef]
11. M. L. Rynes, D. A. Surinach, S. Linn, M. Laroque, V. Rajendran, J. Dominguez, O. Hadjistamoulou, Z. S. Navabi, L. Ghanbari, G. W. Johnson, M. Nazari, M. H. Mohajerani, and S. B. Kodandaramaiah, “Miniaturized head-mounted microscope for whole-cortex mesoscale imaging in freely behaving mice,” Nat. Methods 18, 417–425 (2021). [CrossRef]
12. W. Zong, R. Wu, S. Chen, J. Wu, H. Wang, Z. Zhao, G. Chen, R. Tu, D. Wu, Y. Hu, Y. Xu, Y. Wang, Z. Duan, H. Wu, Y. Zhang, J. Zhang, A. Wang, L. Chen, and H. Cheng, “Miniature two-photon microscopy for enlarged field-of-view, multi-plane and long-term brain imaging,” Nat. Methods 18, 46–49 (2021). [CrossRef]
13. A. Klioutchnikov, D. J. Wallace, J. Sawinski, K. M. Voit, Y. Groemping, and J. N. D. Kerr, “A three-photon head-mounted microscope for imaging all layers of visual cortex in freely moving mice,” Nat. Methods 20, 610–616 (2022). [CrossRef]
14. H. Guo, Q. Chen, W. Qin, W. Qi, and L. Xi, “Detachable head-mounted photoacoustic microscope in freely moving mice,” Opt. Lett. 46, 6055–6058 (2021). [CrossRef]
15. W. Zong, H. A. Obenhaus, E. R. Skytoen, H. Eneqvist, N. L. de Jong, R. Vale, M. R. Jorge, M. B. Moser, and E. I. Moser, “Large-scale two-photon calcium imaging in freely moving mice,” Cell 185, 1240–1256 (2022). [CrossRef]
16. P. Shekhtmeyster, D. Duarte, E. M. Carey, A. Ngo, G. Gao, J. A. Olmstead, N. A. Nelson, and A. Nimmerjahn, “Trans-segmental imaging in the spinal cord of behaving mice,” Nat. Biotechnol. (2023). [CrossRef]
17. L. Lin and E. Keeler, “Progress of MEMS scanning micromirrors for optical bio-imaging,” Micromachines 6, 1675–1689 (2015). [CrossRef]
18. C. Gorecki and S. Bargiel, “MEMS scanning mirrors for optical coherence tomography,” Photonics 8, 6 (2020). [CrossRef]
19. S. T. S. Holmstrom, U. Baran, and H. Urey, “MEMS laser scanners: a review,” J. Microelectromech. Syst. 23, 259–275 (2014). [CrossRef]
20. D. Wang, C. Watkins, and H. Xie, “MEMS mirrors for LiDAR: a review,” Micromachines 11, 456 (2020). [CrossRef]
21. B. R. Samanta, F. Pardo, T. Salamon, R. Kopf, and M. S. Eggleston, “Low-cost electrothermally actuated MEMS mirrors for high-speed linear raster scanning,” Optica 9, 251–257 (2022). [CrossRef]
22. M. Imboden, J. Chang, C. Pollock, E. Lowell, M. Akbulut, J. Morrison, T. Stark, T. G. Bifano, and D. J. Bishop, “High-speed control of electromechanical transduction: advanced drive techniques for optimized step-and-settle response of MEMS micromirrors,” IEEE Control Syst. 36, 48–76 (2016). [CrossRef]
23. Y. Tan, R. Dong, Z. Li, G. Chai, and M. Li, “Online optimization control with predictive gradient descent for MEMS micro-mirrors,” IEEE Trans. Ind. Electron. 69, 7307–7317 (2022). [CrossRef]
24. T. Liu, T. Pan, P. Wang, S. Qin, and H. Xie, “Scanning optimization of an electrothermally-actuated MEMS mirror for applications in optical coherence tomography endoscopy,” Sens. Actuators A 335, 113377 (2022). [CrossRef]
25. C. Sun, Y. Liu, B. Li, W. Su, M. Luo, G. Du, and Y. Wu, “Modeling and optimization of a novel ScAlN-based MEMS scanning mirror with large static and dynamic two-axis tilting angles,” Sensors 21, 5513 (2021). [CrossRef]
26. W. Zong, R. Wu, M. Li, Y. Hu, Y. Li, J. Li, H. Rong, H. Wu, Y. Xu, Y. Lu, H. Jia, M. Fan, Z. Zhou, Y. Zhang, A. Wang, L. Chen, and H. Cheng, “Fast high-resolution miniature two-photon microscopy for brain imaging in freely behaving mice,” Nat. Methods 14, 713–719 (2017). [CrossRef]
27. A. Klioutchnikov, D. J. Wallace, M. H. Frosz, R. Zeltner, J. Sawinski, V. Pawlak, K. M. Voit, P. St.J. Russell, and J. N. D. Kerr, “Three-photon head-mounted microscope for imaging deep cortical layers in freely moving rats,” Nat. Methods 17, 509–513 (2020). [CrossRef]
28. O. M. Carrasco-Zevallos, C. Viehland, B. Keller, R. P. McNabb, A. N. Kuo, and J. A. Izatt, “Constant linear velocity spiral scanning for near video rate 4D OCT ophthalmic and surgical imaging with isotropic transverse sampling,” Biomed. Opt. Express 9, 5052–5070 (2018). [CrossRef]
29. Q. Smithwick, E. Seibel, P. Reinhall, and J. Vagners, “Control aspects of the single-fiber scanning endoscope,” Proc. SPIE 4253, 176–188 (2001). [CrossRef]
30. D. Y. Kim, K. Hwang, J. Ahn, Y. H. Seo, J. B. Kim, S. Lee, J. H. Yoon, E. Kong, Y. Jeong, S. Jon, P. Kim, and K. H. Jeong, “Lissajous scanning two-photon endomicroscope for in vivo tissue imaging,” Sci. Rep. 9, 3560 (2019). [CrossRef]
31. J. Wang, G. Zhang, and Z. You, “Design rules for dense and rapid Lissajous scanning,” Microsyst. Nanoeng. 6, 101 (2020). [CrossRef]
32. J. Wang, G. Zhang, and Z. You, “Improved sampling scheme for LiDAR in Lissajous scanning mode,” Microsyst. Nanoeng. 8, 64 (2022). [CrossRef]
33. K. Hwang, Y. H. Seo, J. Ahn, P. Kim, and K. H. Jeong, “Frequency selection rule for high definition and high frame rate Lissajous scanning,” Sci. Rep. 7, 14075 (2017). [CrossRef]
34. P. Isola, J.-Y. Zhu, T. Zhou, and A. A. Efros, “Image-to-image translation with conditional adversarial networks,” in Conference on Computer Vision and Pattern Recognition (CVPR) (2017), pp. 5967–5976.
35. M. Mirza and S. Osindero, “Conditional generative adversarial nets,” arXiv, arXiv:1411.1784 (2014). [CrossRef]
36. K. B. Patel, W. Liang, M. J. Casper, V. Voleti, W. Li, A. J. Yagielski, H. T. Zhao, C. Perez Campos, G. S. Lee, J. M. Liu, E. Philipone, A. J. Yoon, K. P. Olive, S. M. Coley, and E. M. C. Hillman, “High-speed light-sheet microscopy for the in-situ acquisition of volumetric histological images of living tissue,” Nat. Biomed. Eng. 6, 569–583 (2022). [CrossRef]
37. W. Qin, W. Qi, T. Jin, H. Guo, and L. Xi, “In vivo oral imaging with integrated portable photoacoustic microscopy and optical coherence tomography,” Appl. Phys. Lett. 111, 263704 (2017). [CrossRef]
