Open Access
Add to BibSonomy
Abstract
Stimulated Raman scattering (SRS) microscopy is increasingly employed for highly specific, label-free, and high-speed bioimaging. Despite its benefits, SRS is susceptible to spurious background signals caused by competing effects, which lower the possible imaging contrast and sensitivity. An efficient approach to suppress these undesired background signals is frequency-modulation (FM) SRS, which exploits the competing effects’ weak spectral dependence compared to the SRS signal’s high spectral specificity. We propose an FM-SRS scheme realized with an acousto-optic tunable filter, which presents a few advantages compared to other solutions presented in the literature. In particular, it can perform automated measurements from the fingerprint to the CH-stretching region of the vibrational spectrum without any manual adjustment of the optical setup. Moreover, it allows simple all-electronic control of the spectral separation and relative intensities of the pair of probed wavenumbers.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Stimulated Raman scattering (SRS) microscopy is a coherent Raman scattering (CRS) technique that exploits molecular vibrations as a contrast mechanism to image samples without perturbing their structure and properties. Thanks to its fast imaging speed (more than 1000-times faster than spontaneous Raman [1] under certain conditions), optical sectioning capability, and immunity to autofluorescence, it has increasingly been utilized in microscopy, in particular for biomedical applications [2–7]. Moreover, compared to the other widely employed CRS technique, i.e., coherent anti-Stokes Raman scattering (CARS), SRS exhibits further advantages such as a linear dependence of the signal on the concentration of molecules of interest and the absence of the so-called nonresonant background (NRB), a spurious contribution created by electronic four-wave mixing process which can considerably reduce the image contrast [8] and distort spectra by interfering with the resonant vibrational signal. Instead, vibrational spectra obtained with SRS are nearly identical to the ones obtained with spontaneous Raman, enabling the exploitation of extensive libraries of molecular spectra already present in literature.
The SRS phenomenon occurs when two frequency-detuned laser beams, a pump beam at frequency $\omega _P$ and a Stokes beam at frequency $\omega _S$ interact with the specimen. If the frequency difference between the two fields matches the frequency of a vibrational mode of a molecule, the molecular transition rate to the vibrational state is enhanced, due to stimulated excitation. This process results in a power loss at the pump frequency (stimulated Raman loss, SRL) and a power gain at the Stokes frequency (stimulated Raman gain, SRG). The signal is usually detected by intensity-modulating one of the two beams and probing the loss or gain on the other beam by demodulating it with a lock-in amplifier (LIA) at the same frequency.
Despite all the advantages that SRS microscopy presents, in the intensity-modulation approach (IM-SRS) the pure Raman signal can be superimposed by spurious background signals caused by a few competing effects. These effects can be classified into two main categories [8–11]: a) nonlinear transient absorption effects — mostly related to the two-photon absorption (TPA) and excited-state absorption (ESA) processes — and b) nonlinear transient scattering effects — including cross-phase modulation (XPM) and thermal lensing (TL). More specifically, the TPA process consists of the quasi-simultaneous absorption of two photons with identical or different frequencies (in the latter case it is referred to as two-color two-photon absorption, TCTPA), exciting a molecule from a lower energy state to a higher energy one. When a pump and a Stokes photons are simultaneously absorbed, the related intensity decrease on both beams is picked up and misinterpreted as an additional Raman signal because of the intensity-modulation technique used in SRS microscopy. A similar phenomenon happens with ESA, which involves the excitation of electronic energy levels but relatively long-lived [12]. On the other hand, nonlinear transient scattering effects generate a local variation of the refractive index in the region of the material where the laser beams are focused, varying the divergence of a probe beam coaxially propagating into the medium. In the case of thermal lensing, when a laser beam propagates into an absorbing medium the photons absorbed in the focal volume increase the temperature locally, causing a refractive index modulation. The XPM effect is a change in the real part of the refractive index experienced by one beam caused by the presence of another beam and is due to the optical Kerr effect [13]. In both TL and XPM, the variation of the probe beam divergence is converted inevitably into a power variation at the detector when there is a finite aperture in the collection path for this beam. When using the conventional IM-SRS scheme, these effects generate a heterodyne background signal that is indistinguishable from the Raman signal and doesn’t carry any chemical information. In standard single-wavenumber IM-SRS measurements this additional background signal can’t be eliminated, due to the fact the spurious effects generating it are quasi-instantaneous, nonlinear, and spatially non-uniform in heterogeneous samples. The typical approaches adopted in IM-SRS setups to reduce the spurious background generated by these competing effects are different. The two-photon absorption process can be reduced using longer wavelengths [14], while the effect of cross-phase modulation is generally reduced using a condenser with a numerical aperture (NA) higher than the one of the focusing objective [15]. However, the extent of reduction might not be sufficient when measuring weak SRS signals, e.g. when the concentration of targeted molecules is extremely low, as it often occurs in biological samples. Moreover, these solutions are not always applicable, for example when performing experiments on live cells: in fact, the preferable option in this type of analysis is using stage-top incubators, which consist of a closed chamber with controlled temperature and humidity, and a transparent area on the top for the beam transmission. The height of the chamber can reach a few centimeters, hence the use of high-NA condensers is hindered because of their very short working distance [16].
Therefore, several solutions, often based on heavily modified setups, have been developed to mitigate these spurious background signals [11,17–19]. In particular, an efficient approach to suppress the undesired background signals is the so-called spectral-modulation or frequency-modulation SRS (FM-SRS) scheme, which exploits the narrowband spectral dependence of the SRS signal, as opposed to the weak spectral dependence of the nonlinear transient scattering and nonlinear transient absorption effects (Fig. 1). In fact, they are almost spectrally flat over tens of nanometers, while the SRS effect, being related to resonant nuclear vibrations, has very sharp spectral features ($\approx$ 1 nm [9]).
Fig. 1. Broad spectral dependence of the background signals due to spurious effects, respectively cross-phase modulation (XPM), two-colour two-photon absorption (TCTPA) and thermal lensing (TL), compared to the narrow spectral dependence of the SRS signal.
By modulating the wavelength of the probe beam on- and off-Raman resonance in a spectral interval of a few nanometers, it is possible to selectively detect the SRS process.
Over the recent years, a variety of SRS setups for background reduction via FM have been reported in the literature [9,20–24]. However, most of the currently implemented FM-SRS solutions present some constraints, either working only in limited spectral regions, needing modifications of the optical setup when performing measurements at different Raman shifts, or presenting a fixed spectral distance between the two selected wavelengths.
We propose a new FM-SRS microscope which surpasses most of the above limitations guaranteeing a broadband background suppression without the need for modification of the optical setup. It is implemented for the first time with an acousto-optic tunable filter (AOTF), exploiting its ability to operate in a broad spectral window and switch rapidly between different on-demand Raman shifts/frequencies. We demonstrate its effective performance in the elimination of spurious background signals of different nature with various samples.
2. Methods
2.1 FM-SRS setup
We implemented the FM-SRS acquisition modality in our existing fingerprint-to-CH-stretch continuously tunable SRS microscopy system [25] (sketched in Fig. 2). In brief, the system is based on a dual-beam femtosecond laser whose fixed-wavelength output acts as the Stokes beam and is intensity-modulated at 5 MHz using an acousto-optic modulator (AOM) and spectrally filtered using a folded 4f spectral shaper. The tunable output of the laser, acting as the pump beam, is dynamically spectrally filtered by an AOTF. A balanced detection scheme is used to suppress the laser excess noise [25]. The two beams are focused on the specimen by using a water immersion objective (Nikon Plan Apo IR SR 60X WI) with 1.27 NA. The power of the transmitted pump beam is measured by a large-area silicon photodiode and the SRS signal is retrieved using a LIA (HF2LI, Zurich Instruments).
Fig. 2. Scheme of the SRS microscope. PBS: polarization beamsplitter; PD: detection photodiode; AOM: acousto-optic modulator; AOTF: acousto-optic tunable filter; LIA: lock-in amplifier.
The AOTF in use (TF950-500-1-5-NT2, Gooch&Housego) is driven by a direct digital synthesizer (DDS) that allows independent analog (intensity) and digital (blanking) control of up to eight channels (spectral filters) (AODS Synth DDS 8 CH, Gooch&Housego) [26]. For each channel, four different frequency profiles—a collection of grouped configuration parameters, such as the frequency and the maximum amplitudes of the acoustic wave in the modulator—can be pre-configured and stored in the DDS’s memory [27]. Two external digital signals (TTL) per channel control the profile’s selection—activating individually each of the four programmed profiles—and an analog signal (0-1 V range) per channel controls the amplitude of its acoustic wave. In our scheme for FM-SRS, the selection of two wavelengths corresponding to the on-resonance and off-resonance Raman shifts required the use of two profiles, out of the four available, of one of the DDS’s channels. The needed analog and digital signals were generated by a multifunction input/output electric board (PXIe 6363 NI-DAQ, National Instruments). For the FM-SRS scheme to obtain the maximum suppression of background signals, the optical power at the two selected wavelengths—filtered by the AOTF from the spectrally broader pump beam (see Fig. 3)—had to be equalized. We achieved this equalization by making the amplitude of the input analog signal to alternate between two different values synchronously with the selected frequency profile. The maximum achievable FM speed—i.e. the frequency ($f_{AOTF}$) at which the pump beam wavelength could be alternated—was estimated experimentally to be 40 kHz, essentially limited by the AOTF performance.
Fig. 3. Overlay of the broad pump beam spectrum (centered at around 793 nm in this case) output from the laser (black line) with the two narrowband spectral components (red line) filtered by the AOTF, before (left) and after (right) the equalization.
In our scheme, the difference in the measured signal’s amplitude at on- and off-resonance wavenumbers results in new frequencies components, corresponding to the sum ($f_{AOM}+f_{AOTF}$), and the difference ($f_{AOM}-f_{AOTF}$) between the AOM and AOTF modulation frequencies. Hence, the differential measurement intrinsic to the FM-SRS approach was readily performed by demodulating the measured signal at both of those frequencies, using the two independent oscillators available in our LIA. Demodulating at both frequencies and adding up the acquired signals increases the signal-to-noise ratio by a factor of $\sqrt {2}$ compared to a single-demodulation-frequency acquisition.
Of note, the use of the AOTF allows the straightforward and fast selection of any pair of wavelengths within the spectral bandwidth (about 150 cm$^{-1}$) of the pump beam for any set central wavelength. Additionally, the center wavelength of the pump beam can be tuned over a very broad spectral range, allowing us to perform FM-SRS measurements from the fingerprint to the CH stretching without any change or manual adjustment of components in our optical setup. This is a significant advantage when compared to other solutions presented in the literature that are limited in the usable spectral range and necessitate manual and time-consuming adjustments in the optical setup [9,20–22].
2.2 Cell culture
Human hepatocellular carcinoma cells (American Type Culture Collection, Hep G2 [HEPG2] ATCC HB-8065) were cultured for 4 days in glucose-free media (A14430, Gibco, Thermo Fisher Scientific) supplemented with 25 mM glucose-d7 (DLM-2062, Cambridge Isotope Laboratories, Inc.), 10% fetal bovine serum, 1% penicillin-streptomycin, and 1% L-glutamine.
3. Results and discussion
The efficiency of our configuration in canceling undesired background signals of diverse nature was tested on different samples, each presenting a different dominant background-causing competing effect.
3.1 Suppression of nonlinear transient absorption background
Dark human hairs are rich in melanin, a pigment with a high TPA cross section [28], therefore they present a high TPA background signal when their Raman spectrum is measured. In order to demonstrate the efficient cancellation of this background with FM-SRS, a brown hair sample was sandwiched between two coverglasses, and the SRS spectrum has been measured (Fig. 4). As expected [9], the hair presents a spectrum with a higher TPA background compared to a reference spectrum from a blond hair, where less melanin is present. The spectra were measured by using a 1.2 NA objective (Zeiss C-Apochromat 63x/1.20 W Corr M27) in collection. The used combination of focusing and collection objectives, with a nearly identical NA, was adopted to minimize the nonlinear transient scattering contribution, allowing to evaluate the contribution of TPA to the background signal only. The pump power at 1660 cm$^{-1}$ was 18 mW and the Stokes power was 21 mW, measured at the sample plane.
Fig. 4. Normalized SRS spectra of a brown (red) and a blond hair (black). The amide I Raman peak at 1660 cm$^{-1}$ in the brown hair sample sits on top of a larger background caused by TPA from the melanin pigment.
In order to show the cancellation of the TPA background signal with our FM-SRS setup, images of the brown hair sample have been collected on-resonance for the amide I band (at 1660 cm$^{-1}$) and off-resonance (at 1730 cm$^{-1}$) using both the single-wavenumber IM-SRS and the FM-SRS modalities. Images of a 40x40 $\mu$m$^{2}$ area were collected on 100x100 pixels and with 0.5 ms integration time per pixel. The images obtained with the IM-SRS configuration, on-resonance at 1660 cm$^{-1}$ (corresponding to the Raman amide I band) and off-resonance at 1730 cm$^{-1}$ (outside of the Raman peak centered at 1660 cm$^{-1}$), are shown in Fig. 5(a,b). With the FM configuration, the images have been collected on-resonance by shifting between 1660 cm$^{-1}$ and 1730 cm$^{-1}$ (Fig. 5(c)), and off-resonance by shifting between 1720 cm$^{-1}$ and 1730 cm$^{-1}$ (Fig. 5(d)). One can notice that the two IM-SRS images look nearly identical, due to the high TPA background effect which dominates over the SRS signal. Conversely, when comparing the two images obtained with the FM-SRS configuration, while on-resonance (Fig. 5(c)) the pure Raman signal of the amide I band is visible, off-resonance (Fig. 5(d)) the hair is no longer visible in the image due to the elimination of the spurious TPA background. These images prove the effective cancellation of TPA parasitic signal in SRS measurements with the developed FM configuration.
Fig. 5. (a-b) Brown hair sample imaged with the IM-SRS configuration, respectively on-resonance (a) at 1660 cm$^{-1}$ and off-resonance (b) at 1730 cm$^{-1}$ (as shown in the spectrum insight at the right bottom of the images). (c-d) Brown hair sample imaged with the FM-SRS configuration; on-resonance (c), shifting between 1660 cm$^{-1}$ and 1730 cm$^{-1}$, and off-resonance (d), alternating between 1720 cm$^{-1}$ and 1730 cm$^{-1}$.
3.2 Suppression of nonlinear transient scattering background
Another set of measurements was performed on human hepatocellular carcinoma cells. The cells were cultured in a medium supplemented with deuterated glucose in order to induce de novo synthesis of deuterated lipids, accumulating in lipid droplets (LDs). The images were captured using two different collection optics: a) a 0.4 NA lens (1" diameter, f= 30 mm, Thorlabs), whose NA is significantly smaller than that of the focusing objective, to greatly increase the background signals caused by transient scattering, to the point of dominating the collected signal; b) a 1.0 NA water dipping objective (Olympus LumPlanFL N 60X WD), to minimize the contribution of the nonlinear transient scattering effects in IM-SRS and compare the results with the background cancellation achieved by the FM-SRS scheme with the low-NA collection. The use of a water dipping objective in collection, having a NA somewhat lower than what used in the previous experiment, was required by the fact that cells were in a coverglass bottom Petri dish.
The IM-SRS images were collected at 2140 cm$^{-1}$, on-resonance for the CD-stretch vibration, and off-resonance at 2050 cm$^{-1}$. The FM-SRS images have been collected modulating between 2050 and 2140 cm$^{-1}$ (on-resonance) and between 2050 and 2060 cm$^{-1}$ (off-resonance). The presence of a spectral baseline due to the background signal caused by nonlinear transient scattering effects is evident by comparing the spectrum captured from one cellular LD using the 0.4 NA lens vs. the high-NA objective in collection (see Fig. 6).
Fig. 6. SRS spectra of a lipid droplet in a hepatocellular carcinoma cell, measured using as a collection optics either a 1.0 NA objective (red line) or a 0.4 NA lens (black line).
The images collected with IM-SRS modality clearly show a high spurious background with low-NA collection (Fig. 7(a,b)) when compared with the ones taken with the high-NA objective (Fig. 7(c,d)). Indeed, the transient scattering effect contribution, higher in the LDs, makes them still very evident in the off-resonance image (Fig. 7(b)). Conversely, the images taken with FM-SRS (Fig. 7(e,f)) clearly show a signal coming only from the LDs when modulating between 2050 and 2140 cm$^{-1}$, and no signal when modulating between the two off-resonance wavenumbers, in a similar way as what happens with high-NA collection in IM-SRS (Fig. 7(c,d)). This suggests the use of a FM-SRS configuration with low-NA collection as an alternative to the need to use high-NA collection objectives in IM-SRS to suppress the background signal caused by transient scattering effects. In fact, a configuration with a high-NA objective in collection is not always applicable, for example when performing experiments on live cells with stage-top incubators [16], whose close chamber is a few centimeters in height, preventing the use of short-working distance objectives.
Fig. 7. Images of a hepatocellular carcinoma cell measured using the IM-SRS and FM-SRS configurations, and different collection optics. First row: images obtained using IM-SRS with a with a low-NA collection lens in collection, (a) on-resonance for the CD stretch vibration (2140 cm$^{-1}$), and (b) off-resonance (2050 cm$^{-1}$). Second row: images taken using IM-SRS with a high-NA objective (1.0 NA) in collection (c) on-resonance for the CD stretch vibration (2140 cm$^{-1}$), and (d) off-resonance (2050 cm$^{-1}$). Third row: images collected using FM-SRS with low NA collection (e) on-resonance, modulating between 2050-2140 cm$^{-1}$, and (f) off-resonance, modulating between 2050-2060 cm$^{-1}$. Scale bars are 10 $\mu$m.
4. Conclusions
We presented a new configuration for the suppression of spurious signals in SRS microscopy. It implements a FM scheme by using a narrowband AOTF that can be rapidly tuned over a very broad spectral range. This configuration offers advantages over other solutions presented in the literature as it doesn’t require any physical modification of our single-wavenumber IM-SRS optical setup, therefore it is possible to seamlessly switch between the IM-SRS and the FM-SRS acquisition modalities. Moreover, the proposed FM-SRS scheme can perform measurements in the whole fingerprint to CH-stretch vibrational spectral range, with a freely adjustable spectral distance between the two selected wavelengths, when they fall within the spectral bandwidth of the pump beam output from the laser. The performance of the proposed configuration was tested by performing SRS imaging on different samples, showing successful cancellation of background signals caused by nonlinear transient scattering and multi-photon absorption processes. Worth to mention, for samples presenting background signals due to transient scattering, our configuration produced imaging results comparable to the gold standard images obtained with a high-NA objective collection. Additionally, these results suggest that using a low-NA collection lens and the FM-SRS method could be an alternative to the use of IM-SRS with high-NA collection objectives. This low-NA configuration is particularly advantageous for live cell experiments for which stage-top incubators are needed.
Funding
King Abdullah University of Science and Technology (OSR-CRG2018-3747).
Disclosures
The authors declare 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 authors upon reasonable request.
References
1. F. Hu, L. Shi, and W. Min, “Biological imaging of chemical bonds by stimulated Raman scattering microscopy,” Nat. Methods 16(9), 830–842 (2019). [CrossRef]
2. C. W. Freudiger, W. Min, B. G. Saar, S. Lu, G. R. Holtom, C. He, J. C. Tsai, J. X. Kang, and X. S. Xie, “Label-free biomedical imaging with high sensitivity by stimulated Raman scattering microscopy,” Science 322(5909), 1857–1861 (2008). [CrossRef]
3. C. H. Camp Jr and M. T. Cicerone, “Chemically sensitive bioimaging with coherent Raman scattering,” Nat. Photonics 9(5), 295–305 (2015). [CrossRef]
4. J.-X. Cheng and X. S. Xie, “Vibrational spectroscopic imaging of living systems: An emerging platform for biology and medicine,” Science 350(6264), aaa8870 (2015). [CrossRef]
5. W. Tipping, M. Lee, A. Serrels, V. Brunton, and A. Hulme, “Stimulated Raman scattering microscopy: an emerging tool for drug discovery,” Chem. Soc. Rev. 45(8), 2075–2089 (2016). [CrossRef]
6. L. Wei, F. Hu, Y. Shen, Z. Chen, Y. Yu, C.-C. Lin, M. C. Wang, and W. Min, “Live-cell imaging of alkyne-tagged small biomolecules by stimulated Raman scattering,” Nat. Methods 11(4), 410–412 (2014). [CrossRef]
7. R. C. Prince and E. O. Potma, “Coherent raman scattering microscopy: capable solution in search of a larger audience,” J. Biomed. Opt. 26(06), 060601 (2021). [CrossRef]
8. P. Berto, E. R. Andresen, and H. Rigneault, “Background-free stimulated raman spectroscopy and microscopy,” Phys. Rev. Lett. 112(5), 053905 (2014). [CrossRef]
9. D. Zhang, M. N. Slipchenko, D. E. Leaird, A. M. Weiner, and J.-X. Cheng, “Spectrally modulated stimulated Raman scattering imaging with an angle-to-wavelength pulse shaper,” Opt. Express 21(11), 13864–13874 (2013). [CrossRef]
10. H. Zhu, H. J. Lee, and D. Zhang, “Background-free stimulated Raman scattering imaging by manipulating photons in the spectral domain,” in Stimulated Raman Scattering Microscopy, (Elsevier, 2022), pp. 137–146.
11. L. Genchi, S. P. Laptenok, and C. Liberale, “Background signals in stimulated Raman scattering microscopy and current solutions to avoid them,” Adv. Phys.: X 8(1), 2176258 (2023). [CrossRef]
12. D. Fu, T. Ye, T. E. Matthews, G. Yurtsever, and W. S. Warren Sr, “Two-color, two-photon, and excited-state absorption microscopy,” J. Biomed. Opt. 12(5), 054004 (2007). [CrossRef]
13. V. Degiorgio and I. Cristiani, “Modulators,” in Photonics, (Springer, 2016), pp. 121–144.
14. D. Zhang, M. N. Slipchenko, and J.-X. Cheng, “Highly sensitive vibrational imaging by femtosecond pulse stimulated Raman loss,” J. Phys. Chem. Lett. 2(11), 1248–1253 (2011). [CrossRef]
15. K. Popov, A. Pegoraro, A. Stolow, and L. Ramunno, “Image formation in CARS and SRS: effect of an inhomogeneous nonresonant background medium,” Opt. Lett. 37(4), 473–475 (2012). [CrossRef]
16. A. Bertoncini, S. P. Laptenok, L. Genchi, V. P. Rajamanickam, and C. Liberale, “3D-printed high-NA catadioptric thin lens for suppression of XPM background in Stimulated Raman Scattering microscopy,” J. Biophotonics 14(5), e202000219 (2021). [CrossRef]
17. M. Andreana, M.-A. Houle, D. J. Moffatt, A. Ridsdale, E. Buettner, F. Légaré, and A. Stolow, “Amplitude and polarization modulated hyperspectral stimulated Raman scattering microscopy,” Opt. Express 23(22), 28119–28131 (2015). [CrossRef]
18. A. Lombardini, P. Berto, J. Duboisset, E. R. Andresen, S. Heuke, E. Büttner, I. Rimke, S. Vergnole, V. Shinkar, P. De Bettignies, and H. Rigneault, “Background-suppressed SRS fingerprint imaging with a fully integrated system using a single optical parametric oscillator,” Opt. Express 28(10), 14490–14502 (2020). [CrossRef]
19. S. Heuke, A. Lombardini, E. Büttner, and H. Rigneault, “Simultaneous stimulated Raman gain and loss detection (SRGAL),” Opt. Express 28(20), 29619–29630 (2020). [CrossRef]
20. W. Yang, A. Li, Y. Suo, F.-K. Lu, and X. S. Xie, “Simultaneous two-color stimulated Raman scattering microscopy by adding a fiber amplifier to a 2 ps OPO-based SRS microscope,” Opt. Lett. 42(3), 523–526 (2017). [CrossRef]
21. A. H. Hill, E. Munger, A. T. Francis, B. Manifold, and D. Fu, “Frequency modulation stimulated Raman scattering microscopy through polarization encoding,” J. Phys. Chem. B 123(40), 8397–8404 (2019). [CrossRef]
22. H. Xiong, N. Qian, Z. Zhao, L. Shi, Y. Miao, and W. Min, “Background-free imaging of chemical bonds by a simple and robust frequency-modulated stimulated Raman scattering microscopy,” Opt. Express 28(10), 15663–15677 (2020). [CrossRef]
23. S. Heuke, I. Rimke, B. Sarri, P. Gasecka, R. Appay, L. Legoff, P. Volz, E. Büttner, and H. Rigneault, “Shot-noise limited tunable dual-vibrational frequency stimulated Raman scattering microscopy,” Biomed. Opt. Express 12(12), 7780–7789 (2021). [CrossRef]
24. K. Wallmeier, T. Würthwein, N. Lemberger, M. Brinkmann, T. Hellwig, and C. Fallnich, “Frequency modulation stimulated Raman scattering scheme for real-time background correction with a single light source,” Biomed. Opt. Express 14(1), 315–325 (2023). [CrossRef]
25. S. P. Laptenok, V. P. Rajamanickam, L. Genchi, T. Monfort, Y. Lee, I. I. Patel, A. Bertoncini, and C. Liberale, “Fingerprint-to-CH stretch continuously tunable high spectral resolution stimulated Raman scattering microscope,” J. Biophotonics 12(9), e201900028 (2019). [CrossRef]
26. L. Genchi, A. Bucci, S. P. Laptenok, A. Giammona, and C. Liberale, “Hadamard-transform spectral acquisition with an acousto-optic tunable filter in a broadband stimulated Raman scattering microscope,” Opt. Express 29(2), 2378–2386 (2021). [CrossRef]
27. Crystal Technology, Inc., AOTF Controllers and FSK Operation (2010).
28. T. Ye, G. Yurtsever, M. Fischer, J. D. Simon, and W. S. Warren, “Imaging melanin by two-photon absorption microscopy,” in Multiphoton Microscopy in the Biomedical Sciences VI, vol. 6089 (SPIE, 2006), pp. 332–338.
