Open Access
Add to BibSonomy
Abstract
Visible light optical coherence tomography (OCT) is an emerging non-invasive imaging modality that offers new opportunities for anatomical and functional imaging of biological tissues. Time-stretch dispersive Fourier transform, also known as photonic time-stretch, is an all-optical processing method that enables real-time Fourier transformation of ultrafast optical signals and allows for OCT at high A-scan rates. In this work, a working prototype of a photonic time-stretch OCT (TS-OCT) method in the visible wavelength region is proposed and experimentally demonstrated. The proposed visible-light TS-OCT system achieves unprecedented throughput of 100 giga voxels/second and OCT volume rate of 4,000 volumes/second and can be used to expand the range of applications of TS-OCT systems.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Optical coherence tomography (OCT) is a volumetric imaging technique that is used to capture images from optical scattering media [1,2]. OCT has been employed in many applications including in-vivo analysis of biological tissues [3–8], non-destructive material inspection [9–11], and applications in ophthalmology, dermatology, and dentistry [1,12,13]. Several methods in the past decades have been developed to implement OCT. This includes time-domain OCT [14], spectral discrimination by swept-source OCT [15], and spectral discrimination by Fourier-domain OCT [16]. In the Fourier domain OCT method, a segment of the electromagnetic spectrum containing fringes in which the interference has occurred is analyzed using an optical spectrum analyzer [16]. Several contributions have been made in the Fourier domain OCT method to improve the axial scan rates as high as 90 MHz [17].
Time-stretch dispersive Fourier transform (TS-DFT) is one of the successful methods that have been developed to improve the scanning speed of Fourier domain OCT methods [18–20]. In TS-DFT method, a highly dispersive medium, usually a long section of optical fiber, is used to map the spectral components of the input optical signal at the output so the optical spectrum can be captured in real-time and single-shot in the time domain using a high-speed oscilloscope [18]. This concept is called time-stretch dispersive Fourier transform [18,19], real-time Fourier transformation [20], or simply photonic time-stretch. Photonic time-stretch techniques can benefit from distributed Raman amplification to boost the signal to noise ratio [19]. Depending on the target application, multiple groups have developed TS-OCT systems at different wavelengths including 0.8 µm [17] and 1.5 µm [21,22] wavelengths.
There have been multitude of developments in the OCT field toward visible wavelength. The main motivations for visible-wavelength OCT are twofold: (i) shorter wavelengths provide better depth resolution and (ii) in some applications visible wavelength provides superior tissue scattering and absorption contrast [23–26]. For example, visible wavelength OCT has been shown to be effective in molecular imaging [23], visualizing skin microvasculature for dermatology applications [24], retinal oximetry [25] and microvascular hemoglobin mapping [26], to name a few. However, implementing TS-OCT in the visible wavelength range has been challenging due to the lack of suitable source laser, dispersive media with low loss and high-dispersion factor, and optical elements that can handle visible wavelength light efficiently.
In this work, a time-stretch OCT technique is proposed and experimentally demonstrated that operates in the visible wavelength range. Main challenges that have been overcome include using a dispersive medium with low loss and high-dispersion factor, utilizing optical elements that can handle visible wavelength light effectively, applying proper optical filtering to avoid distortions in the time-stretch signal, and nonlinear phase distortion compensation. The implemented visible light TS-OCT system can provide high-speed volumetric imaging in some applications that visible wavelength OCT is beneficial.
2. Principle of operation
In optical domain, Michelson interferometry system involves the use of a beam splitter and two mirrors [27]. This system can be used to estimate the spatial distance difference between the two mirrors by analyzing the interference pattern in the reflected signal. In OCT systems, one of the mirrors is replaced by the sample under test [1]. Figure 1 illustrates a basic configuration of a Fourier domain OCT system based on Michelson interferometer.
However, conventional Fourier domain OCT systems suffer from low scan rates due to the limited sweep rate of the optical spectrum analyzer. Despite the impressive capabilities of conventional Fourier domain OCT systems, they are often hindered by relatively low scan rates. This limitation largely stems from the optical spectrum analyzer's restricted sweep rate. To circumvent this bottleneck, the time-stretch dispersive Fourier transform (TS-DFT) method has been proposed and successfully implemented [18–20,28–30]. TS-DFT capitalizes on a section of a temporally dispersive medium to map the spectral response of an ultrashort optical signal into the time domain. This mapping enables real-time capture of the signal spectrum using an oscilloscope, as depicted in Fig. 2. The dispersive medium plays a crucial role in this process by ‘stretching’ the output temporal signal and ‘slowing it down.’ The result is a time-stretched pulse that carries the spectral interference pattern in the time domain. In this method, the frequency information of the signal is transformed into temporal information, effectively converting a high-speed frequency measurement problem into a slower-speed time measurement problem. This transformation allows us to leverage the high-speed capabilities of electronic oscilloscopes to perform real-time detection and analysis of the spectral interference patterns. This approach offers significant advantages in high-speed imaging applications, enabling the capture of dynamic processes in real time with high resolution. The TS-DFT has found widespread use in various applications, including high-speed imaging, real-time spectroscopy, and high-speed digitization, underlining its versatile utility and potential for further advancements in the field [18].
Fig. 2. Illustration of frequency to time mapping phenomenon for a short pulse passing through time-stretch dispersive Fourier transform (TS-DFT) system.
Photonic time-stretch OCT (TS-OCT) combines the TS-DFT method with spectral interferometry to implement real-time complex-field spectral interferometry [28], as shown in Fig. 3. In this method, interference signal composed of reference and sample signals are passed through a high-dispersive medium to map the resulting spectral interference to the time domain. The spectral interference can be then captured using a high-speed oscilloscope in real-time and single-shot.
Fig. 3. Schematic of the Fourier-domain OCT system based on photonic time-stretch technology. The system employs a highly dispersive medium to perform time-stretch dispersive Fourier transform (TS-DFT) on the spectral interference signal, enabling high-speed spectral-domain imaging.
In this work, a successful implementation of the TS-OCT method in the visible wavelength is presented. This work could open new opportunities for real-time volumetric imaging in biomedical imaging applications where visible wavelength provides better depth resolution and scattering or absorption contrast.
3. Proposed system description
The proposed system description is shown in Fig. 4 and outlines the experimental setup used for proof-of-concept demonstrations. A source ultrafast pulsed laser was used in this setup, specifically the NKT SuperK EXTREME, which had a maximum output power of 6W and emitted light at wavelengths between 450 nm and 2400 nm. The repetition rate of the laser was set to 2.686 MHz. To limit the input laser wavelength range to the visible range, a short-pass optical filter (Thorlabs, FES0750) with a cutoff wavelength of 750 nm was incorporated. A collimator with a numerical aperture (NA) of 0.56 was used to couple the supercontinuum laser filtered light to a 50/50 fiber-based coupler (Thorlabs model TW670R5A2, fiber type 630 HP).
Fig. 4. Experimental setup for proof-of-concept demonstrations. The ultrafast pulsed laser emits light at wavelengths between 450 nm and 2400 nm, which is limited to the visible range using a short-pass optical filter. The filtered light is coupled to a 50/50 fiber-based coupler. The reference arm includes an inline polarization controller and an adjustable time delay stage. The sample arm includes 2D galvo scanners and a scanning lens to collimate the beams on the sample. The reflected light is launched to a section of single-mode fiber for TS-DFT and detection using a photoreceiver and an oscilloscope.
In the reference arm, an inline polarization controller was employed to maximize the fringe visibility in the spectral interference waveform. An adjustable time delay stage was used in the reference arm to tune the delay between the reference arm and the sample arm. In the sample arm, a 2D galvo scanner was used to scan the sample under test in the x-y directions. Each galvo mirror could sweep the operating angle range at the speed of up to 2 KHz. Two synched arbitrary waveform generators (AWGs) were used to control the speed and pattern of each galvo mirror. A pulse generator was used to trigger the two AWGs and was also used to trigger the oscilloscope. A scanning lens was used to collimate the beams on the sample. The galvo scanner and scanning assembly were mounted on a 3D printed custom housing installed on an imaging stand. To scan the x-y space by using 100 by 100 points, one of the galvo mirrors was set to operate at 1 KHz and the other galvo mirror was set to operate at 10 Hz. The power on the sample was measured to be approximately 9.8 mW.
The reflected light from the coupler was directed into a 1-km segment of single-mode fiber (SMF-28), inducing a substantial dispersion of -356.25 ps/nm on the interference signal within the visible wavelength region. The dispersive fiber subsequently carried out a time-stretch dispersive Fourier transform (TS-DFT) on the interference signal. This process generated the spectral interference in the time domain through frequency-to-time mapping. The temporal width of the input pulse was approximately 5 ps, and after undergoing filtering and time-stretch, the pulse duration expanded to roughly 57 ns. Given the dispersion coefficient value of -356.25 ps/nm and the input pulse duration of 5 ps, our system was operating within the far-field regime of the time-stretch concept, as determined by the far-field criteria in time-stretch literature [29,30]. Moreover, to circumvent the overlap of neighboring pulses post time-stretch, we ensured that the time-stretch pulse duration was less than the 372.3 ns laser period.
We employed a 38 GHz high-speed photoreceiver (Newport 1474-A), followed by a wideband RF amplifier, to detect the consecutive spectral interference patterns. This particular photoreceiver was chosen because it covers lower wavelengths in the visible range down to 590 nm with an appropriate optical response. The consecutive spectral interference patterns were subsequently detected using a 20 GHz and 100 GS/s real-time oscilloscope. The system could potentially use any photoreceiver that covers visible wavelengths and has an RF bandwidth wider than 20 GHz. Oscilloscope data was streamed to a computer using high-speed ethernet communication. The received data was post-processed on the computer to visualize 3D OCT captured images.
The measured optical spectrum at the output of the dispersion fiber using a scanning grating-based optical spectrum analyzer (Yokogawa AQ6370D) is shown in Fig. 5 with the blue dotted line. For this measurement, the reference arm was blocked, and a mirror was used as the sample in the sample arm to focus on the shape of the spectrum itself. The captured TS-DFT waveform on the oscilloscope for the same setup, followed by frequency to time mapping scaling, is also shown in Fig. 5 with the green dashed line. To ensure accurate absolute wavelength measurements in TS-DFT, the wavelength axis was calculated using the absolute wavelength method based on calculating the absolute delay between the measured TS-DFT signal and a reference signal from the laser captured using another channel on the oscilloscope. The attenuation in the lower wavelengths in the TS-DFT signal is primarily caused by the lower responsivity of the photoreceiver in this wavelength region. The calculated TS-DFT spectral trace after calibration of the amplitude roll-off is shown in Fig. 5 with the red solid line. The results shown in Fig. 5 demonstrate the promise of the proposed setup to work within the 590 nm to 750 nm visible wavelength range, corresponding to a bandwidth of 160 nm.
Fig. 5. Blue dotted line: Measured optical spectrum at the output of the dispersion fiber using an optical spectrum analyzer with the reference arm blocked and a mirror used as sample in the sample arm. Green dashed line: Captured TS-DFT waveform on the oscilloscope followed by frequency to time mapping scaling. Red solid line: Calculated TS-DFT spectral trace after calibration of the amplitude roll-off.
In the choice of spectral analysis equipment, we employed the Yokogawa AQ6370D for its dynamic range of 110 dB, which significantly improved our signal-to-noise ratio. However, for applications requiring lower wavelength coverage, alternative models such as the Yokogawa AQ6374, capable of extending coverage down to 350 nm, may be more appropriate despite its slightly reduced dynamic range of 100 dB. The selection of the spectrum analyzer should be guided by the specific demands of the system under study, balancing the requirements for wavelength coverage and signal sensitivity.
Nonlinear dispersion in the employed dispersive medium causes phase distortions in the measured spectral interference. The effect of phase distortion on the measured spectral interference is shown in Fig. 6(a). For this experiment, a mirror was used in the reference arm and a mirror was used in the sample arm to create a single delay for the spectral interference. The calculated phase distortion in radians using Hilbert transform method is shown in Fig. 6(a) with the red dashed line. The phase distortion was calibrated digitally using the Hilbert transform method. The calibrated spectral interference pattern after phase distortion correction is shown in Fig. 6(b).
Fig. 6. (a) Effect of phase distortion on the measured spectral interference with calculated phase distortion in radian shown in red dashed line. (b) Calibrated spectral interference pattern after phase distortion correction.
Figure 6(c) presents a comparison between the OCT traces before and after the implementation of phase distortion correction in the Fourier domain. In this particular experiment, the spectral interference is associated with a single delay, leading to an expectation of a single peak in the Fourier domain waveform. Prior to phase compensation, the peak appears to be obscured within the Fourier domain. However, upon applying phase compensation, the single peak in the Fourier domain becomes distinctly visible.
To ensure the robustness of our imaging capabilities, an analysis of the signal-to-noise ratio (SNR) of the OCT images generated by this system was conducted. The SNR was determined as the ratio of the peak signal level to the noise level within the Fourier domain, as illustrated in Fig. 6(c). To acquire these measurements, a mirror was placed in the sample arm, providing a controlled environment for determining the SNR and power on sample was 9.8 mW. After the implementation of phase compensation, the system's SNR was found to be 21 dB, measured from the peak to the noise floor in the Fourier domain. While this value is satisfactory for certain applications, we acknowledge that further enhancements can be made. Potential strategies under consideration include the use of averaging techniques, the implementation of balanced detection, the incorporation of more sensitive detectors, and the use of digitizers with higher effective number of bits (ENOBs).
4. Experimental results
In this section, experimental results of the implemented time-stretch OCT method in the visible wavelength is presented. In the first test, we used a USAF 1951 Resolution negative Test Target (shown in Fig. 7(a)) with a thickness of 1 mm. The sample slide was placed on a mirror to also capture reflections from the bottom level of the slide with better signal-to-noise ratio. The scan area for the probe is indicated in Fig. 7(a) with a yellow solid box. The proposed method successfully captured a high-resolution 3D OCT image of the resolution target sample in both x-y plane and in z direction, as shown in Fig. 7(b). The 3D OCT image has 100 by 100 pixels in the x-y plane and 80 pixels in the z direction. We selected depth cross-sections at the top and bottom surfaces of the slide to demonstrate the method's capability in capturing the sample's details in x-y planes at two different depths (1 mm art). We would like to highlight that no smoothing or interpolation has been used in these results, and the resolution target image was captured without any post-processing. As demonstrated in Fig. 7, the proposed method was successful in capturing the 3D image of the resolution target sample, showcasing its potential for 3D visible wavelength imaging using photonic time-stretch OCT method.
Fig. 7. Experimental results of the proposed visible-wavelength time-stretch OCT method. (a) A USAF 1951 Resolution negative Test Target with a thickness of 1 mm was used as a sample. The scan area for the probe is indicated by a yellow solid box. (b) 3D OCT image of the resolution target sample captured using the proposed method in both x-y plane and in z direction. Depth cross-sections at the top and bottom surfaces of the slide are shown to demonstrate the method's capability in capturing sample details. (c) Line cross-section of OCT image in the y-direction. The location of the line is indicated in the (b) with a red arrow.
To assess the lateral resolution of the proposed time-stretch OCT method, we analyzed the captured 3D OCT images of the USAF 1951 Resolution negative Test Target. Figure 7(c) shows the line cross-section of OCT image in the y-direction. The location of the line is indicated in the Fig. 7(b) with a red arrow. The resolution was calculated as the smallest distinguishable feature size in the line image, which was found to be 30 µm in the y direction.
In the second experimental demonstration, the objective was to investigate the Fabry Perot cavity of a virtually imaged phased array (VIPA) device using the proposed time-stretch OCT method. The aim was to determine the distance between the two sides of the cavity across the x-y plane. We also used this experiment to estimate the axial resolution of the proposed OCT method. To accomplish this, we employed a setup that consisted of the proposed OCT system with a spectral range of 590-750 nm and a VIPA device with a cavity size of 112${\pm} $. 2 µm.
Figure 8(a) displays two images of the VIPA device from different angles. The top picture shows the x-y scan area used in the experiment, while the bottom picture highlights the location of the Fabry Perot cavity. In Fig. 8(b), we present the OCT image obtained using the proposed experimental setup, with two cross-sections corresponding to the two sides of the cavity. Figure 8(c) shows the line cross-section of OCT image in the z-direction. The location of the line is indicated in the Fig. 8(b) with a red arrow. The distance between the two cavity walls is estimated by calculating the distance between the two peaks in the Fig. 8(c). The OCT image estimates that the distance between the two cavity sides of the VIPA device is 109.8 µm, demonstrating the capability of the proposed method to accurately measure the dimensions of the target object.
Fig. 8. (a) Images of the VIPA device captured from two perspectives. The top image illustrates the x-y scan area employed for the experimental demonstration, while the bottom image highlights the location of the Fabry-Perot cavity. (b) OCT image of the VIPA device acquired using the proposed experimental setup, featuring two cross-sections representing two sides of the cavity. (c) Axial line cross-section of the OCT image in the z-direction, with its location marked by a red arrow in image (b). The OCT line image estimates the distance between the two cavity sides of the VIPA device to be 109.8 µm.
The axial resolution of the proposed OCT method is influenced by several limiting factors. These include free-space optical components, such as the numerical aperture (NA) and aberrations of the scanning lens, the bandwidth of the source laser, optimal dispersion calibration, and amplitude roll-off correction. Each of these factors contributes to the overall performance and resolution of the OCT system, making their optimization crucial for obtaining accurate and high-quality imaging results. In order to determine the axial resolution of the proposed OCT method, we examined the line image depicted in Fig. 8(c). Axial resolution was calculated based on the smallest distinguishable feature size present in the line image. Our analysis revealed that the axial resolution was 13.2 µm in the z direction, representing the minimum resolvable feature size using this OCT technique.
Despite SMF-28's cutoff lying above the visible range, its use in this experiment was justified by its combination of a considerable dispersion factor and low insertion loss, factors that are integral to SNR-focused applications such as this application. However, the potential for multi-mode operation within the visible range could affect this system's dispersion characteristics and the spectral resolution.
The axial resolution in the proposed system is affected by several factors. Multi-modal dispersion complicates the complete compensation for nonlinear dispersion and residuals, potentially limiting the attainable resolution. Additionally, the ENOB of the high-speed digitizer used in our experiment plays a role in defining the axial resolution. High-speed digitizers, such as the one used in our setup, often have a lower ENOB compared to low bandwidth digitizers, presenting an inherent constraint on the system's performance. Improving the signal-to-noise ratio within the system can be also helpful to enhance the OCT system axial resolution.
In the final experimental demonstration, we assessed the capacity of the proposed time-stretch OCT method to image a moving target, specifically a rotating DC motor shaft head. Figure 9(a) displays the sample under examination in the left image, while the right image showcases the disassembled components of the DC motor shaft, highlighting three identifiable parts with top surfaces in the DC motor shaft head, as indicated by three red arrows and numbers. Line imaging in the x direction was employed in this experiment, with the y-dimension generated by the sample moving beneath the OCT line imaging. The location of the line imaging cross line is illustrated by a red solid line in the left image of Fig. 9(a), and the rotation direction of the DC motor is denoted by a black arrow.
Fig. 9. Experimental results of the proposed time-stretch OCT method for imaging a rotating DC motor shaft head using line imaging. (a) Left image displays the sample under test, with the red solid line representing the location of the OCT line imaging cross line and the black arrow denoting the direction of rotation. Right image exhibits the disassembled components of the DC motor shaft, with three identifiable parts and their top surfaces indicated by 3 red arrows and numbers. (b) OCT volumetric image depicting three distinguishable surfaces of the DC motor shaft head. (c) Top view of the three surfaces. This volumetric OCT image was acquired in 250 µs at a throughput of 100 Giga Voxels per second and a volume rate of 4,000 volumes per second.
Figure 9(b) presents the resulting OCT volumetric image of the rotating DC motor shaft head, revealing the x-y planes at the three distinguishable surfaces. The top view of these surfaces is depicted in Fig. 9(c). This experimental demonstration shows the potential for the proposed method to be applied in the analysis of dynamic moving objects. Due to the short duration of the laser pulses used in the time-stretch OCT method (typically <1 ns), the blurriness effect caused by sample movement is minimal.
The OCT line imager consisted of 100 pixels in each line in the x direction, with 240 useful pixels out of 500 raw pixels at each depth and 500 pixels in the y direction. The OCT volume size in this experiment was 12 mega voxels (100 × 500 × 240 voxels), and the OCT image was captured at a throughput of 100 giga voxels per second, yielding an OCT volume frame rate of approximately 4,000 volumes per second. Each OCT volume was acquired in 250 µs, with the total number of OCT volumes captured in a single shot limited to four volumes by the oscilloscope's record length (100 million samples). Nevertheless, high-speed real-time oscilloscopes with record lengths as high as 2 giga samples are currently available on the market, potentially increasing the total number of OCT frames captured at 100 giga voxels per second by at least one order of magnitude.
In the experiments presented here, the relatively high optical power of 9.8 mW applied to the sample is acknowledged. Moreover, the importance of sensitivity, which is the minimum sample reflectivity required to achieve an SNR of 1, is recognized. In this case, a minimum detectable reflected power from the sample of about 6 µW was achieved without using any post optical amplification (e.g. Raman amplification), RF amplification, signal SNR boost using balanced detection schemes or digital enhancing such as averaging. Potential strategies to enhance system sensitivity and SNR include the use of optical amplification (e.g. Raman amplification), RF amplification, averaging techniques, the implementation of balanced detection, the incorporation of more sensitive detectors, and the use of digitizers with higher ENOBs. These enhancements aim to improve the sensitivity and versatility of this high-speed imaging system while maintaining a careful balance between the applied optical power and the resulting image quality.
The potential applications of the time-stretch OCT instrument presented in this paper are vast, given its high-speed imaging capability. In the realm of biomedical imaging, this instrument can be utilized for the rapid, non-invasive imaging of biological tissues, potentially facilitating early disease diagnosis. The technology may also be employed in material science for real-time imaging of material surfaces at the microscopic level, a significant advantage in industries requiring stringent quality control. Industrial inspection could benefit from the instrument's ability to monitor fast-moving parts or processes in real-time, aiding in rapid identification of anomalies or defects. Moreover, with the integration of deep learning techniques, the instrument could be employed for label-free cell classification, providing a means for real-time analysis of large cell populations. The capability to observe fast dynamic phenomena in real-time would be a considerable advantage in disciplines such as physics or engineering. Furthermore, the technology could be applied to real-time spectroscopy, enabling rapid, high-resolution spectral analysis in a variety of scientific and industrial applications.
5. Conclusion
In conclusion, we have proposed and demonstrated a time-stretch OCT method capable of performing 3D imaging of targets in the visible-wavelength range. The proposed method has proven effective in imaging both stationary and dynamic samples with high resolution and minimal motion artifacts. With an unprecedented imaging throughput as high as 100 giga voxels per second and update rate as high as 4,000 volumes per second, the proposed method significantly outperforms conventional OCT systems, making it a promising tool for high-speed imaging applications in areas such as biomedical imaging, material science, and industrial inspection.
The experimental results indicate that the proposed method successfully imaged a USAF 1951 Resolution negative Test Target, a virtually imaged phased array (VIPA) device, and a rotating DC motor shaft head. The method accurately determined the distance between the two sides of the Fabry Perot cavity within the VIPA device. Furthermore, the high frame rate of the proposed method enabled the imaging of dynamic samples, such as the rotating DC motor shaft head, with minimal motion artifacts.
Expanding further on the potential of real-time imaging, the ability to capture high-resolution images at such high speed opens up new possibilities in many domains. For example, in biomedical imaging, the real-time aspect could allow for the detection of rapid physiological changes or the tracking of fast cellular processes. In industrial inspection, the high-speed imaging could facilitate more efficient and comprehensive quality control processes. In material science, this method could provide new ways to study dynamic processes and material transformations.
However, it is acknowledged that there are challenges and limitations inherent in this field. The speed of real-time imaging is not solely determined by the imaging technology itself, but also by the computational power available for processing and interpreting the collected data. Advancements in faster oscilloscopes and improved computational capabilities, such as end-to-end GPU or FPGA processing, will undoubtedly bolster the potential of real-time imaging. Moreover, sensitivity remains a pivotal issue that warrants further investigation in these systems. Enhancing the sensitivity of this system will be a key area of focus in our future work, as it is critical for expanding the scope and applicability of our method in various domains.
Disclosures
The author declares no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the author upon reasonable request.
References
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]
2. J. M. Schmitt, “Optical coherence tomography (OCT): a review,” IEEE J. Select. Topics Quantum Electron. 5(4), 1205–1215 (1999). [CrossRef]
3. A. F. Fercher, K. Mengedoht, and W. Werner, “Eye-length measurement by interferometry with partially coherent light,” Opt. Lett. 13(3), 186–188 (1988). [CrossRef]
4. W. Drexler and J. G. Fujimoto, Optical Coherence Tomography: technology and applications (biological and medical physics, biomedical engineering), (Springer, 2008).
5. M. J. Suter, S. K. Nadkarni, G. Weisz, A. Tanaka, F. A. Jaffer, B. E. Bouma, and G. J. Tearney, “Intravascular optical imaging technology for investigating the coronary artery,” JACC Card. Imag. 4(9), 1022–1039 (2011). [CrossRef]
6. B. J. Vakoc, R. M. Lanning, J. A. Tyrrell, T. P. Padera, L. A. Bartlett, T. Stylianopoulos, L. L. Munn, G. J. Tearney, D. Fukumura, R. K. Jain, and B. E. Bouma, “Three-dimensional microscopy of the tumor microenvironment in vivo using optical frequency domain imaging,” Nat. Med. 15(10), 1219–1223 (2009). [CrossRef]
7. D. C. Adler, Y. Chen, R. Huber, J. Schmitt, J. Connolly, and J. G. Fujimoto, “Three-dimensional endomicroscopy using optical coherence tomography,” Nat. Photonics 1(12), 709–716 (2007). [CrossRef]
8. B. J. Vakoc, D. Fukumura, R. K. Jain, and B. E. Bouma, “Cancer imaging by optical coherence tomography: preclinical progress and clinical potential,” Nat. Rev. Cancer 12(5), 363–368 (2012). [CrossRef]
9. A. Nemeth, G. Hannesschläger, E. Leiss-Holzinger, K. Wiesauer, and M. Leitner, “Optical coherence tomography – applications in non- destructive testing and evaluation,” Opt. Coh. Tomog. Masanori Kawasaki, DOI: 10.5772/53960, 2013.
10. H. S. G. Costa, O. R. P. Bellon, and L. Silva, “Biometric identification with 3D fingerprints acquired through optical coherence tomography,” presented at Conf. on graphics, patterns and images (SIBGRAPI), São José dos Campos, 2016.
11. F. Zechel, R. Kunze, N. König, and R. H. Schmitt, “Optical coherence tomography for non-destructive testing,” Tech. Mess. 87(6), 404–413 (2020). [CrossRef]
12. J. Welzel, “Optical coherence tomography in dermatology: a review,” Skin Res Technol. 7(1), 1–9 (2001). [CrossRef]
13. A. Baumgartner, S. Dichtl, C. K. Hitzenberger, H. Sattmann, B. Robl, A. Moritz, A. F. Fercher, and W. Sperr, “Polarization-sensitive optical coherence tomography of dental structures,” Caries Res. 34(1), 59–69 (2000). [CrossRef]
14. Y. Watanabe, F. Sajima, T. Itagaki, K. Watanabe, and Y. Shuto, “High-speed linear detection time domain optical coherence tomography with reflective grating-generated spatial reference delay,” Appl. Opt. 48(18), 3401–3406 (2009). [CrossRef]
15. 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]
16. A. Fercher, C. K. Hitzenberger, G. Kamp, and S. Y. El-Zaiat, “Measurement of intraocular distances by backscattering spectral interferometry,” Opt. Commun. 117(1-2), 43–48 (1995). [CrossRef]
17. G. Keisuke, A. Fard, O. Malik, F. Gilbert, A. Quach, and B. Jalali, “High-throughput optical coherence tomography at 800 nm,” Opt. Express 20(18), 19612–19617 (2012). [CrossRef]
18. A. Mahjoubfar, D. Churkin, D. S. Barland, N. Broderick, S. K. Turitsyn, and B. Jalali, “Time stretch and its applications,” Nat. Photonics 11(6), 341–351 (2017). [CrossRef]
19. K. Goda, D. R. Solli, K. K. Tsia, and B. Jalali, “Theory of amplified dispersive Fourier transformation,” Phys. Rev. A 80(4), 043821 (2009). [CrossRef]
20. J. Azana, L. R. Chen, M. A. Muriel, and P. W. E. Smith, “Experimental demonstration of real-time Fourier transformation using linearly chirped fibre Bragg gratings,” Electron. Lett. 35(25), 2223–2224 (1999). [CrossRef]
21. J. Xu, X. Wei, L. Yu, C. Zhang, J. Xu, K. K. Y. Wong, and K. K. Tsia, “High-performance multi-megahertz optical coherence tomography based on amplified optical time-stretch,” Biomed. Opt. Express 6(4), 1340–1350 (2015). [CrossRef]
22. S. Moon and D. Y. Kim, “Ultra-high-speed optical coherence tomography with a stretched pulse supercontinuum source,” Opt. Express 14(24), 11575–11584 (2006). [CrossRef]
23. F. E. Robles, C. Wilson, G. Grant, and A. Wax, “Molecular imaging true-colour spectroscopic optical coherence tomography,” Nat. Photonics 5(12), 744–747 (2011). [CrossRef]
24. Xiao Shu, Lisa Jane Beckmann, and Hao F. Zhang, “Visible-light optical coherence tomography: a review,” J. Biomed. Opt. 22(12), 121707 (2017). [CrossRef]
25. J. Yi, Q. Wei, W. Liu, V. Backman, and H. F. Zhang, “Visible-light optical coherence tomography for retinal oximetry,” Opt. Lett. 38(11), 1796–1798 (2013). [CrossRef]
26. S. P. Chong, C. W Merkle, C. Leahy, H. Radhakrishnan, and V. J. Srinivasan, “Quantitative microvascular hemoglobin mapping using visible light spectroscopic optical coherence tomography,” Biomed. Opt. Express 6(4), 1429–1450 (2015). [CrossRef]
27. L. Lepetit, G. Chériaux, and M. Joffre, “Linear technique of phase measurement by femtosecond spectral interferometry for applications in spectroscopy,” J. Opt. Soc. Am. B 12(12), 2467–2474 (1995). [CrossRef]
28. H. Asghari, Y. Park, and J. Azaña, “Complex-field measurement of ultrafast dynamic optical waveforms based on real-time spectral interferometry,” Opt. Express 18(16), 16526–16538 (2010). [CrossRef]
29. J. Azaña and M. A. Muriel, “Real-time optical spectrum analysis based on the time-space duality in chirped fiber gratings,” IEEE J. Quantum Electron. 36(5), 517–526 (2000). [CrossRef]
30. Y. Xiao, Y. Meng, T. Hao, N. Shi, W. Li, and M. Li, “Ultra-fast full-field optical characterization of CW lasers based on optical frequency comb, wavelength-to-time mapping and phase-diversity,” Opt. Express 29(24), 39874–39884 (2021). [CrossRef]
