Open Access
Add to BibSonomy
Abstract
We propose polarization-sensitive second-harmonic generation microscopy based on off-axis interferometry (OI-PSHG) by recording the complex field of a wide-field second-harmonic generation (SHG) image and performing polarization measurements. With the ability to record the SHG signals associated with different positions simultaneously, the proposed method exhibits a higher imaging frame rate than raster scanning-based SHG microscopy. The molecular orientation (in terms of their symmetric axis) of tendon collagen fibrils and myosin in muscle is resolved in three dimensions from a subset of polarization-resolved SHG holograms. With the present configuration, it takes approximately 0.01 s to acquire an image with 128 × 128 pixels, which is mainly limited by the excitation power density for wide-field illumination. For the same data throughput using pixel-by-pixel scanning, 0.16-s-long acquisition is required, with the pixel dwell time of 10 µs. Offering the ability to perform wide-field imaging and polarization measurements, the present work lays the foundation for fast SHG microscopy using complex deconvolution and harmonic tomography.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Second-harmonic generation (SHG) microscopy is a powerful nonlinear imaging modality that can reveal the structure of cells and biological tissues without exogenous labeling [1–3]. SHG microscopy is based on a second-order nonlinear effect (i.e., the χ(2) effect) that depends quadratically on the excitation intensity, thus enabling intrinsic optical sectioning. Although SHG microscopy has made great strides along with the introduction of novel laser sources, detectors, and improved computational capability, image acquisition in state-of-the-art SHG microscopy utilizes pixel-wise raster scanning using a tightly focused laser beam, which constrains the imaging frame rate. The most promising solution of this issue is wide-field SHG imaging, in which an ensemble of sampled points is illuminated and imaged using an area array camera. Wide-field SHG microscopy has been reported [4–6], in which a sample is placed at the defocused position of a condenser lens with a low numerical aperture (NA), yielding wide-field illumination. Weak SHG signals are recorded using an electron multiplier charge-coupled (EMCCD) camera. Unfortunately, wide-field SHG microscopy can only detect the intensity of a non-center symmetric sample but not its phase. To overcome this limitation, interferometry combined with SHG microscopy was adopted in this study. In the proposed interferometric SHG (ISHG) approach (Fig. 1(A)), an intense laser pulse with frequency ω is delivered to a set of samples, generating second harmonics and emitting scattered signals at both the fundamental (ω) and second (2ω) harmonic frequencies. A bandpass filter centered at 2ω transmits only the 2ω signal of the samples to the detector. A reference beam at a frequency of 2ω also arrives at the detector; this beam is generated independently using the same pump laser pulse, resulting in the formation of a hologram. Therefore, the coherent characteristic of the harmonic signal allows one to record its complex field from the interferogram, including both the amplitude and phase distributions, using common holography methods [7,8].
Fig. 1. Working principle of the OI-PSHG microscope. (A) Optical layout of the OI-PSHG microscope. BBO: beta barium borate crystal; DM: dichroic mirror; HWP: half-wave plate; Con.: condenser; Obj.: microscope objective; OP: optical filter; BS: beam splitter; P: polarizer; sCMOS: scientific camera with complementary metal-oxide-semiconductor detector. (B) Illustration of the polarizations of the fundamental (${E^{({{\omega_0}} )}}$) and the SHG wave (${E^{({2{\omega_0}} )}}$), orientations of myosin fibril and polarization analyzer (polarizer). Zχ is the symmetric axis of the myosin fibril. (C) Illustration of inverse diffraction from an SHG hologram. (D) Comparison between measured and calculated results using back-propagation of the systematic CPSF, where (D1) and (D2) show the off-axis interferogram and its spatial spectrum, respectively; (D3), (D5) and (D6) show the amplitudes in the xy, xz, and yz planes; (D4), (D7) and (D8) show the corresponding phases. Scale bars: (D1) and (D3) - (D8) 1 µm; (D2) 20 rad/µm.
The interpretation of the SHG contrast is not a trivial task because it results from a combination of local second-order nonlinear properties (local symmetry and organization) and the orientation of the microstructure with respect to both the direction of propagation and polarization of the excitation beam [9,10]. However, solving the inverse problem of SHG polarization, that is, revealing the intrinsic nonlinear structure and symmetry of harmonophores from measured polarization states, has been extensively studied [11]. Previous studies using polarization-sensitive SHG (PSHG) microscopy have explored the symmetrical axis orientation of collagen fibers [12–15], myosin filaments [16], and cartilage [17]. Information about the orientation field allows further characterization of the anisotropic properties of biological tissues that affect their mechanical and rheological properties [18]. In these studies, the orientation of each scanning point was determined by continuous modulation [13] or by the selection of several designed states [19,20] of the incident polarization, and orientation maps were then formed by raster scanning. However, the extra measurements of different polarization states, combined with the laser scanning-based operation mode, increase the overall analysis time and limit the applicability of PSHG microscopy.
Previous studies on polarization-sensitive SHG microscopy typically relied on pixel scanning of modestly focused spots, formed by objectives with relatively low NA, to avoid substantially altering the polarization state in the focus. For instance, an objective with NA = 0.42 was adopted in Ref. [12], resulting in a transverse resolution of approximately 1.5 µm, and axial resolution of approximately 10 µm. In the present work, wide-field SHG microscopy adopted a low-NA condenser for the second-harmonic wave excitation and a relatively high-NA objective working on the imaging configuration to reallocate the SHG signals associated with different positions.
In this paper, we present a scheme and results of off-axis interferometry-based PSHG microscopy (OI-PSHG) in a wide-field imaging mode that aims to improve the measurement speed of PSHG. Fast polarization measurement of three-dimensional (3D) nonlinear structures was achieved by (1) wide-field illumination and imaging instead of pixel-by-pixel scanning utilized by conventional multiphoton microscopes; (2) orientation retraction algorithm based on analyzing only 16 images of characteristic polarization states [20] instead of the conventional continuous polarization modulation method [13]; and (3) fast z-scanning accomplished by reconstructing a 3D nonlinear structure from the measured complex optical field. It is noteworthy that wide-field SHG imaging allows for single-shot complex field measurements by off-axis interferometry instead of the phase-shifting method, thus easing the requirement of environmental stability [21,22]. Measuring the phase distribution of the SHG field also provides the ability to detect molecular polarity according to previous interferometric SHG microscopy based on laser scanning [23–25]. Because the polarities of the non-centrosymmetric components are opposite to each other, the emitted SHG signals are π phase-shifted, owing to the opposite signs of the achiral components of their second order nonlinear susceptibility tensor χ(2). In the present work, holographically recording the complex field revealed this property in a single-shot manner, improving the analysis speed.
2. Materials and methods
2.1 Experimental setup
Wide-field SHG images were acquired with different excitation and detection polarizations using an OI-PSHG microscope (Fig. 1(A)). A femtosecond (fs) pulse laser (Mira HP, Coherent Corp., USA) was tuned to 800 nm and focused onto a beta barium borate (BBO) (Huate Material, China) crystal to generate the SHG signal as the reference for the following off-axis interferometry studies. A collimated laser beam, consisting of the fundamental and SHG waves, was separated by a dichroic mirror (Chroma, T425lpxr), after which the fundamental wave was steered to the sample and the SHG signal to the reference path. The imaged sample was placed near the focus of an aspheric condenser (ACL2018U-B, Thorlabs, USA) with NA = 0.52 for wide-field illumination. The defocus amount was carefully adjusted to achieve a trade-off between the illumination spot size and power density. The 1/e radius of the illuminated spot, measured through the objective, was 32 µm. An objective (Nikon Plan Fluor 40x, NA = 0.75) was used to collect the generated SHG signal, after passing through a short pass filter (Chroma, ET525sp-2p) and a tube lens, to a scientific complementary metal-oxide-semiconductor (sCMOS) camera (Marana, ANDOR). A time-delay configuration was set to the reference beam path to optimize the temporal overlap between the reference and object pulse chains. The reference and object beams were combined using a non-polarized beam splitter (NPBS). To produce a clean reference beam, a spatial filter with a 75-µm-diameter pinhole (P75HK Thorlabs) (not shown in Fig. 1(A)) was placed within the reference path immediately before the NPBS. Three elements were utilized to determine the excitation and generated polarizations: a half-wave plate at the fundamental wavelength (HWP@800 nm) controlled the excitation polarization, a polarizer placed after the NPBS determined the detected polarization, and a half-wave plate at the SHG wavelength (HWP@400 nm) adapted the polarization of the reference beam in the direction of the polarizer. The power reached by the sample was 980 mW. Even though the power of 980 mW seems too high for point scanning microscopy, it was not critical in the proposed setup, because the energy was distributed throughout a volume that was much larger than that of a tight focused point. The 1/e width of the radius of the illuminated area in our case was approximately 32 µm, which was approximately 50 times larger than that in normal point scanning microscopy. This means that the volume of the focusing spot was approximately 503 times larger. Therefore, the energy density was so low that we had to increase the exposure time to the order of milliseconds compared with tens of microseconds for point scanning microscopy [27].
2.2 OI-PSHG image analysis
A standard inverse diffraction algorithm [5] was used to reconstruct the SHG intensity distributions at different depths, as illustrated in Fig. 1(C). In this section, we briefly introduce inverse diffraction for recovering an object’s image, the diffraction field of which is typically measured on a plane. In the Fresnel approximation, inverse diffraction results can be obtained directly. In the algorithm, the observation and measurement planes are assumed to be at $\textrm{z} = {z_0}$ and $\textrm{z} = {z_r}$, respectively. The diffraction propagation distance between the two planes is ${z_{0r}}$. The medium is assumed to be homogeneous under normal circumstances. The challenge is to determine the field at $\textrm{z} = {z_0}$, when it is known at $\textrm{z} = {z_r}$. The key equations are
We used a sample of 95-nm-size ZnO nanoparticles to test the feasibility of the inverse diffraction method of our system. It was reported earlier that ZnO nanoparticles yielded detectable second-harmonic signals owing to the electromagnetic wave confinement by the Mie resonance [26]. Because the sizes of the ZnO nanoparticles are well below the resolution of our objective (Nikon plan fluor 40x, NA = 0.75), an axial scanning and holographic recording of each image would result in a 3D complex point spread function (CPSF). However, if the inverse diffraction algorithm is correct, we should be able to calculate the 3D CPSF by back-propagating a hologram at a defocused position. The results are shown in Fig. 1(D). Figs. 1(D1) and 1(D2) show the off-axis interferogram and the spatial spectrum, respectively. The high-intensity Airy disk resulted in a high-contrast fringe within the dashed circle. The upper band of the spatial spectrum within the dashed circle was isolated and moved to the origin of the spatial domain to complete analytic signal generation. Figs. 1(D3) and 1(D4) show the measured amplitude and phase at the focus, respectively. To measure the 3D CPSF, we scanned 24 steps axially, with a resolution of 0.424 µm. The backpropagated field from the 10th step was also calculated using the first interferogram. The center of the isolated region in the spatial spectrum should be correctly identified; otherwise, a tilted phase ramp should be added to the constructed CPSF. In Figs. 1(D5)–1(D8), the fields above the dashed lines were measured directly, while the fields below the dashed lines were calculated using the inverse diffraction algorithm. This clearly shows good consistency between the directly measured and calculated results.
2.3 Retraction of myosin orientation
Previous studies have shown that SHG in myosin [12,16] and collagen [14] exhibits cylindrical symmetry (C∞) with respect to the local second-order nonlinear susceptibility ${\mathrm{\chi }^{(2 )}}({\boldsymbol r} )$. For any material with cylindrical symmetry, the polarization vector can be written concisely as [13]
Under the above conditions, the SHG intensity in Fig. 1(B) without a polarizer can be obtained by substituting Eq. (4) into Eq. (3) [13], yielding
Because the orientation angle is estimated using intensity ratios, the local concentration of harmonophore c(r) is canceled out, resulting in an algorithm that is insensitive to c(r).
3. Results
We first tested the polarization measurement ability of OI-PSHG using collagen fibrils from the rat tail tendon, which is a common test target owing to its well-ordered and easily identifiable collagen structure, as shown in Fig. 2. The tendon fibril was manually V-shaped (Fig. 2(A)) so that one could observe the molecular orientation change with the tendon trend. We tested the dependence of the overall intensity U on the excitation polarization and found good agreement with the theoretical prediction (Eq. (5), assuming cylinder asymmetry in the tendon fibril) as shown in Fig. S2 (Supplement 1). Using the aforementioned algorithm, the symmetrical axial orientation of the tendon fibril was obtained. In this experiment, we analyzed the orientation angles at various positions of the collagen fibril by translating the sample holder. Figs. 2(B) and 2(C) show two typical results that overlay the SHG overall intensity and orientation vector (quivers) of the axial orientation. In general, the symmetrical axis is parallel to the tendon fibril (Figs. 2(B) and 2(C)), which is consistent with previous results obtained using scanning polarization-sensitive SHG [12].
Fig. 2. Polarization measurement results of collagen fibrils from rat tail tendon. (A) Bright field image. (B) and (C): SHG intensity and polarization vector plots corresponding to ROI 1 and ROI 2 highlighted in (A). (D) Histograms of the orientation angle corresponding to ROI 1 and ROI 2. The dashed lines indicate Gaussian fits. PDF: probability density function. Scale bars: (A) 100 µm; (B) and (C) 5 µm.
The histograms in Fig. 2(D) quantitatively show the orientation angle distributions for regions of interest (ROIs) 1 and 2. The angles that corresponded to relatively high SHG intensities (≥10% of the maximum) were counted to avoid bias from data with a low signal-to-noise ratio (SNR). The Gaussian fits show that the mean values of the orientation angle distributions $\phi $ are 0.16π and 0.25π radians, with standard deviations of 0.23π and 0.24π for ROI 1 and ROI 2, respectively.
We then used an OI-PSHG microscope to measure the filaments of the mouse leg muscle. Figure 3 shows the intensities for different combinations of the excitation and analyzing polarizations, for a 50-µm-thick mice muscle slice. Each row corresponds to the same excitation polarization, and the analyzing angles are 0°, 45°, 90°, and 135°. The data in the same column have the same analyzing angles. The separation between sarcomeres, which appear as cross-striated bands owing to the densely packed myosin/actin filaments inside [1], is clearly seen, and the measured length is approximately 2 µm.
Fig. 3. Wide-field SHG images corresponding to various excitation and detection polarizations. Red and blue arrows denote the excitation and detection polarization directions, respectively. The polarization of both the excitation and SHG waves are 0°, 45°, 90°, and 135°. Scale bar: 5 µm.
Interestingly, the maximal SHG intensity was not obtained when the excitation oscillated parallel to the muscle filaments (perpendicular to the sarcomeres). Indeed, Eq. (5) and Fig. S2 suggest that the maximum should occur when the excitation is polarized at a given angle with respect to the filaments. It also shows that under the maximal excitation condition, the generated second-harmonic signal is polarized more linearly than in parallel to the filaments, as can be seen by comparing the first and second rows in Fig. 3.
The polarization measurement results are shown in Fig. 4. Figures 4(A) and 4(B) show the overall intensity and phase distributions, respectively. No significant phase difference was observed for the different excitation and detection polarizations. The obtained molecular orientations shown in Fig. 4(C) indicate that the symmetrical axes are, in general, perpendicular to the direction of myosin, which agrees well with a previous report [9].
Fig. 4. Polarization measurements of a muscle tissue section of rat. (A) Overall intensities, for different values of the excitation angle U. (B) Map of the SHG phase distribution. Units: radians. (C) Overlay between the orientation vectors and intensity U. (D) Amplified view of the dashed box area in (C). Scale bars: 5 µm in (A)-(C); 1 µm in (D).
To reconstruct the 3D volume image, each hologram with a given combination of excitation and analyzing polarizations was diffracted inversely to the target imaging plane. Fig. 5(A) shows the 3D rendering of the reconstructed muscle tissue section of the rat. Sarcomeres can be clearly seen, organized in a spatial pattern with a period of approximately 1.4 µm, which agrees well with previous reports [7,8]. The depth dependence of the polarization is shown in Fig. 5(C), corresponding to the positions highlighted in Fig. 5(B). Evidently, the polarization direction changed slightly, which may be owing to fibril misalignment.
Fig. 5. 3D reconstruction results and polarization measurements of a muscle tissue section of a rat (thickness, 50 µm). (A) 3D rendering image. (B) Overlay between the orientation vectors and intensity U (different locations with different markers). (C) Depth dependence of the polarization direction. Scale bar: 2 µm.
The penetration depth of approximately 100 µm in Fig. 5(A) was owing to the thickness of the sample, which was only 100 µm. In fact, thicker samples could also be imaged using our polarization-measurement system. However, the objective of this work was to further harmonic tomography research, and the results for thicker samples are not shown herein. This is because harmonic tomography requires weak scattering for relatively thin samples.
4. Discussion and conclusions
Using the present configuration, the time for recording a 128 × 128 pixel image was approximately 0.01 s, mainly limited by the excitation power density resulting from wide-field illumination. For the same data throughput using pixel-by-pixel scanning, 0.16 seconds were required for a pixel dwell time of 10 µs. To determine the molecular orientation angle, 16 images corresponding to different excitation and detection angles were recorded. It took approximately 100 µs for inverse diffraction to a single plane on our computer (CPU, Intel Core i7-10700 K) using MATLAB software (MathWorks Corp., US). The main drawback of the present wide-field illumination SHGM was its relatively low axial resolution, as can be observed from the CPSF in Fig. 1(D). The out-of-focus signal coherently interfered with the on-focus signal. A complex deconvolution algorithm [21] and further harmonic-calculated tomography based on angular scanning of the illumination beam [22], which are both based on wide-field illumination and imaging, can be applied to improve the axial resolution. On the other hand, the present work lays the foundation for polarization-resolved SHG microscopy using the above techniques, which are further developed in our laboratory.
In summary, we have demonstrated a polarization-based SHG holographic microscope, which we used together with sample orientation and susceptibility ratios in 3D on unlabeled biological samples from a set of excitation and analyzer polarization-resolved SHG hologram images. From these data, we numerically reconstructed 3D images of the total (not modulated by the incident field polarization) second-harmonic intensity scattered by the specimens. Given its ability to capture 3D images rapidly, this novel polarization-sensitive SHG holography opens new vistas to high-speed dynamic 3D imaging for studying biological processes, such as muscle mechanics with millisecond temporal resolution.
Funding
Shenzhen International Cooperation Project (GJHZ20190822095420249); National Key Research and Shenzhen Key Projects (JCYJ20200109105404067); National Natural Science Foundation of China (61835009, 61935012, 62127819, 62175163, 62225505); Development Program of China (2017YFA0700402).
Acknowledgement
We thank Meixia Wang for providing the ZnO nanoparticles in the calibration procedure in our imaging system and Prof. Kebin Shi for helpful discussions.
Disclosures
The authors declare no conflicts of interest.
Data availability
The 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.
Supplemental document
See Supplement 1 for supporting content.
References
1. P. J. Campagnola and L. M. Loew, “Second-harmonic imaging microscopy for visualizing biomolecular arrays in cells, tissues and organisms,” Nat. Biotechnol. 21(11), 1356–1360 (2003). [CrossRef]
2. P. Campagnola, “Second harmonic generation imaging microscopy: Applications to diseases diagnostics,” Anal. Chem. 83(9), 3224–3231 (2011). [CrossRef]
3. B. Shen, J. Yan, S. Wang, F. Zhou, Y. Zhao, R. Hu, J. Qu, and L. Liu, “Label-free whole-colony imaging and metabolic analysis of metastatic pancreatic cancer by an autoregulating flexible optical system,” Theranostics. 10(4), 1849–1860 (2020). [CrossRef]
4. O. Masihzadeh, P. Schlup, and R. A. Bartels, “Label-free second harmonic generation holographic microscopy of biological specimens,” Opt. Express 18(10), 9840 (2010). [CrossRef]
5. C. L. Hsieh, R. Grange, Y. Pu, and D. Psaltis, “Three-dimensional harmonic holographic microcopy using nanoparticles as probes for cell imaging,” Opt. Express 17(4), 2880 (2009). [CrossRef]
6. Y. Pu, M. Centurion, and D. Psaltis, “Harmonic holography: a new holographic principle,” Appl. Opt. 47(4), A103–A110 (2008). [CrossRef]
7. U. Schnars and W. Jüptner, “Digital recording and numerical reconstruction of holograms,” Meas. Sci. Technol. 13(9), R85–R101 (2002). [CrossRef]
8. W. Yu, X. Li, R. Hu, J. Qu, and L. Liu, “Full-field measurement of complex objects illuminated by an ultrashort pulse laser using delay-line sweeping off-axis interferometry,” Opt. Lett. 46(12), 2803–2806 (2021). [CrossRef]
9. S. W. Chu, S. Y. Chen, G. W. Chern, T. H. Tsai, Y. C. Chen, B. L. Lin, and C. K. Sun, “Studies of x((2))/x((3)) tensors in submicron-scaled bio-tissues by polarization harmonics optical microscopy,” Biophys. J. 86(6), 3914–3922 (2004). [CrossRef]
10. K. R. Campbell, B. Wen, E. M. Shelton, R. Swader, B. L. Cox, K. Eliceiri, and P. J. Campagnola, “3D second harmonic generation imaging tomography by multi-view excitation,” Optica 4(10), 1171 (2017). [CrossRef]
11. X. Y. Dow, E. L. DeWalt, J. A. Newman, C. M. Dettmar, and G. J. Simpson, “Unified theory for polarization analysis in second harmonic and sum frequency microscopy,” Biophys. J. 111(7), 1553–1568 (2016). [CrossRef]
12. P. Stoller, P. M. Celliers, K. M. Reiser, and A. M. Rubenchik, “Quantitative second-harmonic generation microscopy in collagen,” Appl. Opt. 42(25), 5209–5219 (2003). [CrossRef]
13. P. Stoller, K. Reiser, P. Celliers, and A. Rubenchik, “Polarization-modulated second harmonic generation in collagen,” Biophys. J. 82(6), 3330–3342 (2002). [CrossRef]
14. M. L. Noorlander, P. Melis, A. Jonker, and C. J. F. Van Noorden, “A quantitative method to determine the orientation of collagen fibers in the dermis,” J. Histochem. Cyto 50(11), 1469–1474 (2002). [CrossRef]
15. T. Yasui, K. Sasaki, Y. Tohno, and T. Araki, “Tomographic imaging of collagen fiber orientation in human tissue using depth-resolved polarimetry of second-harmonic generation light,” Opt. Quantum Electron. 37(13-15), 1397–1408 (2005). [CrossRef]
16. S. V. Plotnikov, A. C. Millard, P. J. Campagnola, and W. A. Mohler, “Characterization of the myosin-based source for second-harmonic generation from muscle sarcomeres,” Biophys. J. 90(2), 693–703 (2006). [CrossRef]
17. S. J. Matcher, “A review of some recent developments in polarization-sensitive optical imaging techniques for the study of articular cartilage,” J. Appl. Phys. 105(10), 102041 (2009). [CrossRef]
18. C. Verdier, “Rheological properties of living materials. From cells to tissues,” J. Theo. Med. 5(2), 67–91 (2003). [CrossRef]
19. C. Odin, T. Guilbert, A. Alkilani, O. P. Boryskina, V. Fleury, and Y. Le Grand, “Collagen and myosin characterization by orientation field second harmonic microscopy,” Opt. Express 16(20), 16151–16165 (2008). [CrossRef]
20. C. Odin, Y. Le Grand, A. Renault, L. Gailhouste, and G. Baffet, “Orientation fields of nonlinear biological fibrils by second harmonic generation microscopy,” J. Microsc. 229(1), 32–38 (2008). [CrossRef]
21. Z. Wang, D. Marks, P. Carney, L. Millet, M. Gillette, A. Mihi, P. Braun, Z. Shen, S. Prasanth, and G. Popescu, “Spatial light interference tomography (SLIT),” Opt. Express 19(21), 19907–19918 (2011). [CrossRef]
22. C. Hu, J. J. Field, V. Kelkar, B. Chiang, K. Wernsing, K. C. Toussaint, R. A. Bartels, and G. Popescu, “Harmonic optical tomography of nonlinear structures,” Nat. Photonics 14(9), 564–569 (2020). [CrossRef]
23. S. Bancelin, C. Couture, K. Légaré, M. Pinsard, M. Rivard, C. Brown, and F. Légaré, “Fast interferometric second harmonic generation microscopy,” Biomed. Opt. Express 7(2), 399–408 (2016). [CrossRef]
24. C. Couture, S. Bancelin, J. Van der Kolk, K. Popov, M. Rivard, K. Légaré, G. Martel, H. Richard, C. Brown, S. Laverty, L. Ramunno, and F. Légaré, “The impact of collagen fibril polarity on second harmonic generation microscopy,” Biophys. J. 109(12), 2501–2510 (2015). [CrossRef]
25. S. Bancelin, C. Couture, M. Pinsard, M. Rivard, P. Drapeau, and F. Légaré, “Probing microtubules polarity in mitotic spindles in situ using interferometric second harmonic generation microscopy,” Sci. Rep. 7(1), 6758 (2017). [CrossRef]
26. L. Bonacina, P. Brevet, M. Finazzi, and M. Celebrano, “Harmonic generation at the nanoscale,” J. Appl. Phys. 127(23), 230901 (2020). [CrossRef]
27. A. Dementjev, R. Rudys, R. Karpicz, and D. Rutkauskas, “Optimization of wide-field second-harmonic generation microscopy for fast imaging of large sample areas in biological tissues,” Lith. J. Phys. 60(3), 4301 (2020). [CrossRef]
