Open Access
Add to BibSonomy
Abstract
A supercontinuum (SC) light source enables multispectral photoacoustic imaging at excitation wavelengths in the visible-to-near-infrared range. However, for such a broad optical wavelength range, chromatic aberration is non-negligible. We developed a multispectral optical-resolution photoacoustic microscopy (MS-OR-PAM) setup with a nanosecond pulsed SC light source and a reflective objective lens to avoid chromatic aberration. Chromatic aberrations generated by reflective and conventional objective lenses were compared, and the images acquired using the reflective objective were not affected by chromatic aberration. Hence, MS-OR-PAM with the reflective objective was used to distinguish red blood cells from melanoma cells via spectral subtraction processing.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. INTRODUCTION
Optical-resolution photoacoustic microscopy (OR-PAM) is a microscopic imaging technique based on the photoacoustic (PA) effect. In the PA effect, an optical absorber that absorbs pulsed light produces ultrasound via a thermoelastic expansion process. Because the measured ultrasound waves, called PA signals, have amplitudes that are proportional to the optical energies absorbed by the optical absorbers, OR-PAM images possess optical-absorption contrast. Hemoglobin is one of the dominant optical absorbers in biological tissues at visible wavelengths; hence, OR-PAM is sensitive to hemoglobin in the blood and can be used to image microvasculature, including tumor and retinal vasculature [1]. By selecting appropriate excitation wavelengths, OR-PAM can also be used to image other optical absorbers. For example, cell nuclei have been imaged using UV light [2], melanin was imaged using red to near-infrared (NIR) light [3,4], and adipose tissues have been imaged using NIR light [5]. Multispectral OR-PAM (MS-OR-PAM) is a technique used to image multiple optical absorbers by acquiring OR-PAM images at multiple excitation wavelengths and performing spectral processing, including spectral subtraction. Some groups have reported the use of MS-OR-PAM to separately image oxy-hemoglobin and deoxy-hemoglobin; in such cases, a calculation of the oxygen saturation of the red blood cells (RBCs) to analyze oxygen metabolism was possible [6,7]. The detection of melanoma cells inside capillaries, without interference from RBC signals, to identify circulating tumor cells (CTCs), which are an indicator of skin cancer metastasis, has also been achieved using this imaging modality [3,8].
Fig. 1. Schematic of experimental setup. SC, super continuum; PCF, photonic crystal fiber; RC, reflective collimator; RBE, reflective beam expander; CF, cold filter; BPF, bandpass filter; NDF, neutral density filter; BS, beam splitter; OL, objective lens; TL, tube lens, HPF; high-pass filter.
Supercontinuum (SC) light sources that can produce pulsed light with single-mode beam characteristics and broadband spectral bandwidth from the visible to NIR have been utilized for various microscopic imaging techniques, including stimulated emission depletion microscopy [9], coherent anti-Stokes Raman scattering microscopy [10], and multiphoton microscopy [11]. Recently, nanosecond-pulsed super continuum light sources with sufficiently strong pulse energy for OR-PAM applications have been developed [12,13], and some groups have utilized them as an excitation light source for MS-OR-PAM [13–17]. In MS-OR-PAM, depending on the scanning area and density to acquire a three-dimensional image, point-by-point serial scanning accompanied by more than tens of thousands of light pulses is required. Thus, repetition frequencies of tens of kHz are required to achieve frame rates higher than 1 fps, which is important for in vivo imaging to avoid motion artifacts and capture fast biological events. In addition, each light pulse should have pulse energies greater than tens of nanojoules to produce detectable PA signals. Because there are few wavelength-tunable light sources that satisfy these requirements, some groups have used tunable dye lasers [6,18] or optical parametric oscillators [19,20] with low repetition frequency to acquire MS-OR-PAM, and other groups used multiple high repetition lasers to acquire MS-OR-PAM images [7]. Nanosecond-pulsed SC light sources satisfy the pulse energy and repetition frequency requirements, enabling excitation wavelength tuning from visible to NIR wavelengths using bandpass filters. Thus, a nanosecond-pulsed SC light source is a good light source for MS-OR-PAM. Recently, stimulated Raman fiber laser light sources that produce light pulses with multiple wavelength peaks were used as the excitation light source for MS-OR-PAM [21]. Although the light source can produce light with a relatively narrow linewidth, which facilitates high spectral resolution imaging at multiple wavelengths, a nanosecond pulsed SC light source can produce light with greater spectral bandwidth. Thus, depending on the application, a nanosecond pulsed SC light source may be advantageous for MS-OR-PAM.
To perform MS-OR-PAM from the visible to NIR wavelength range, the problem of chromatic aberration, which mainly results from the wavelength-dependent refractive indices of lens materials and results in wavelength-dependent focal length shifting, should be addressed. In the case of limited spectral bandwidth, chromatic aberration may be negligible when using a conventional achromatic objective lens that compensates for low-level chromatic aberration [21]. However, when tuning the excitation wavelength from the visible to the NIR, chromatic aberration is non-negligible even if an achromatic objective lens is used [22]. For this reason, although thus far SC light sources exist that can produce pulsed light over a wide spectral bandwidth range from visible to NIR, MS-OR-PAM has been performed using only either visible or NIR excitation wavelengths.
In this study, to enable MS-OR-PAM over a wide spectral range from the visible to the NIR, we developed a chromatic aberration-free MS-OR-PAM system with an SC light source and a reflective objective lens. The reflective objective lens, which was based on the Schwarzschild system, ensured chromatic aberration-free optical focusing over a wide spectral band. Although some groups have reported the use of a reflective objective lens for MS-OR-PAM, in those cases the repetition frequency or wavelength tunability of the light sources limited the performance of their systems as shown in Table 1 [19,23]. A ray-tracing simulation study validated our system design concept. We then compared the effect of chromatic aberration on MS-OR-PAM images by switching between the reflective objective lens and a standard refractive objective lens. Furthermore, we performed MS-OR-PAM of RBCs and melanoma cells to demonstrate that the MS-OR-PAM images acquired using the chromatic aberration-free system can be used for spectral processing to distinguish among multiple optical absorbers in a single sample.
Table 1. Comparison of Light Sources Used for MS-OR-PAM with Reflective Objectives
2. METHODS
A. Experimental Setup
A schematic of the MS-OR-PAM system, which was based on a microscope system (CUS-BF, SigmaKoki, Tokyo, Japan), is shown in Fig. 1. Broadband light pulses were produced using a light source (SM30-W, Leukos, Limoges, France). The light pulses had a pulse width of 1 ns and repetition frequency of 27.62 kHz. The light beam was collimated using a reflective collimator (RC08APC-P01, Thorlabs, Newton, NJ) and then expanded using a reflective beam expander (BE02R/M1, Thorlabs, Newton, NJ). The light pulses were filtered using a cold filter (SC1101, Asahi Spectra, Tokyo, Japan) and one of a pair of bandpass filters mounted on a filter wheel (FW2A, Thorlabs, Newton, NJ). The central wavelengths of the light pulses were selected by rotating the filter wheel. The bandpass filters (86-951 and 84-789, Edmund Optics, Barrington, NJ) had passband widths of 50 nm and central wavelengths of 525 nm and 800 nm, respectively. The filtered light pulses were attenuated by a tunable neutral density filter (NDHN-U100, SigmaKoki, Tokyo, Japan) and then used as excitation light pulses for the OR-PAM system. The excitation light pulse was reflected by a beam splitter (BSW10, Thorlabs, Newton, NJ), and then focused by a ${10} \times$ reflective objective (OBLR-10A, SigmaKoki, Tokyo, Japan) with an effective focal length (EFL) of 19.9 mm or a ${10} \times$ standard glass objective lens (EPL-10, SigmaKoki, Tokyo, Japan) with an EFL of 20 mm. The beam diameters of the Gaussian beams entering the objectives were measured to be 6.58 mm and 8.48 mm at 525 nm and 800 nm, respectively, using a beam profiler (SP620, Ophir, Jerusalem, Israel) equipped with a ${4} \times$ beam reducer (SPZ17017, Ophir, Jerusalem, Israel). The difference between the values for the two wavelengths was a result of the wavelength dependence of the numerical aperture of the photonic crystal fiber in the SC light source.
A sample placed on the focal plane absorbed excitation light pulses and produced PA signals. The PA signals were measured using a PA sensor (HD30-3.5-8, HONDA ELECTRONICS, Aichi, Japan) with a central frequency of 31.4 MHz, an element diameter of 3.5 mm, and a focal length of 8.2 mm. The PA signals were amplified by 92 dB using two 46 dB amplifiers (SA-230F5, NF Corporation, Kanagawa, Japan). The amplified PA signals were filtered using a high-pass filter (EF509, Thorlabs, Newton, NJ) to remove low-frequency electrical noise, and were recorded using a digital oscilloscope (PXIe-5164, National Instruments, Austin, TX) operated at a sampling frequency of 500 MHz. The recorded PA signals were enveloped using a Hilbert transform [24]. The imaging target was placed in a water tank mounted on three perpendicularly aligned motorized stages ($x$, OSMS26-50; $y$, HPS80-50X-M5; $z$, SGSP-13ACT-B0; SigmaKoki, Tokyo, Japan). Because the time delay between excitation light irradiation and PA signal detection reflected the distance between the optical absorbers and the PA sensor, the temporal profile of the PA signal measured at a single location reflects the depth profile of the optical absorbers in the imaging target along the PA detection axis (the $z$ axis in our experimental setup). Thus, scanning the imaging target along the $x$ axis generated an $x {-} z$ cross-sectional image, and a raster scan of the imaging target along the $x$ and $y$ axes generated a three-dimensional image. Furthermore, by scanning the imaging target along the $z$ axis, multiple images were acquired while shifting the optical focus of the objective. The digital oscilloscope and motorized stages were controlled by a laboratory-made LabVIEW program. The OR-PAM system also included an LED light source (SLSI-22 W, SigmaKoki, Tokyo, Japan), a tube lens (CU-040, SigmaKoki, Tokyo, Japan), and a charge-coupled device (CCD) camera (CU-171-CA01, SigmaKoki, Tokyo, Japan), which was used to confirm observed sample positions via optical microscopy.
B. Simulation of Light Focusing Using a Reflective Objective
A reflective objective based on the Schwarzschild system was used in this experiment. The Schwarzschild reflective objective is a simple optical system consisting of two spherical mirrors whose centers of curvature coincide [25]. The incident light is first reflected by a smaller spherical mirror (secondary mirror) and then focused by a larger spherical mirror (primary mirror). The focal length, determined by the curvatures of the two spherical mirrors, is independent of the optical wavelength [26,27]. Thus, the reflective objective is known to produce uniform focusing over a wide spectral bandwidth [22]. To confirm the performance of the reflective objective used in this experiment, a ray-tracing simulation of our experimental setup including the reflective objective was performed using Optics Studio (Zemax, Kirkland, WA) in cooperation with the manufacturer of the reflective objective, SigmaKoki Co., Ltd. In the simulation, to evaluate the dependence of the focal length on the wavelength and focusing performance of the reflective objective, Strehl ratios were calculated for multiple planes either side of the focal plane at both 525 nm and 800 nm.
C. Experimental Evaluation of Focus Shift due to Chromatic Aberration
To evaluate the effect of chromatic aberration on the images, a chrome-coated microscope calibration scale (OBJT001, Shibuya Optical, Saitama, Japan) was imaged using the MS-OR-PAM system. A schematic of the experiment is shown in Fig. 2. A hole was made in the bottom of the water tank, and the chrome-coated surface of the calibration scale was attached near the objective lens to minimize the effects of chromic aberration on the sample. The water tank was filled with distilled water. The excitation light illuminated the calibration scale directly, and the immersed PA sensor detected the PA signals produced by the calibration scale through the glass of the calibration scale. To demonstrate the focal shift caused by chromatic aberration, the experiment was performed at excitation wavelengths of 525 nm and 800 nm. The energy of the excitation light was optimized to produce a sufficiently strong PA signal while preventing damage to the sample at 6.2 nJ/pulse.
Fig. 2. Experimental setup to evaluate the focus shift due to chromatic aberrations. By translating the calibration scale in the $z$ axis, the distance between the objective lens and calibration scale was adjusted. The cross-sectional image of the sharp edge of the calibration scale was obtained on each $z$ position by scanning the calibration scale pasted on the bottom of the water tank along the $x$ axis.
To measure the focus shift due to chromatic aberration, cross-sectional images of the edge on the calibration scale were obtained by changing the distance between the calibration scale and objective. Cross-sectional images with different focus positions were obtained by scanning the sample holder along the $x$ and $z$ motorized-stage axes. The sample holder was scanned over a distance of 60 µm with a 0.5 µm step size along the $x$ axis and 80 µm with a 2 µm step size along the $z$ axis.
From the cross-sectional images, the signal intensities along the $x$ axis were calculated as the maximum signal intensities along the $z$ axis. The full width at half-maximum (FWHM) was calculated for each cross-sectional image as a measure of spatial resolution. The FWHM was calculated by fitting the signal intensities along the $x$ axis to an edge spread function. The focal plane at each excitation wavelength was calculated as the distance between the objective and calibration scale corresponding to the narrowest FWHM.
D. MS-OR-PAM Imaging of Melanoma Cells and RBCs
To examine whether spectral processing to distinguish among multiple optical absorbers can be achieved using MS-OR-PAM images acquired from our system with a reflective objective, imaging experiments were performed on RBCs and B16-F0 cells (mouse melanoma). As shown in Fig. 3, oxy-hemoglobin in RBCs strongly absorbs at 525 nm, and melanosomes in melanoma cells strongly absorb at 525 nm and 800 nm. Thus, by using 800 nm excitation light, the melanoma can be imaged with high contrast. In addition, it is expected that a weighted subtraction of the OR-PAM images acquired at 800 nm from those at 525 nm removes the contribution of melanosome and produces high-contrast images of RBCs.
Fig. 3. Absorption spectra of oxy-hemoglobin (15 g/dL) [28] and melanosome [29]. The melanosome used to generate this spectrum was estimated to consist of 30% w/w eumelanin [29] because melanosomes in melanoma cells contain significant amounts of eumelanin [29,30] and melanosomes in B16 melanoma cells contain eumelanin at a concentration of 30% w/w.
First, to test whether the weighted subtraction processing technique can remove the melanosome signal, OR-PAM images of the melanosome cells were acquired at both 525 and 800 nm. B16-F0 cells (mouse melanoma) were grown in Dulbecco’s modified Eagle’s medium containing D-glucose, L-glutamine, and sodium pyruvate (DMEM, 11885–084; Life Technologies, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (SH3091003; Life Technologies, Carlsbad, CA, USA) and 1% antibiotic–antimycotic (15240-062; Life Technologies, Carlsbad, CA, USA) in 5% ${{\rm CO}_2}$. Two hours before the PAM imaging, B16-F0 cells in Hanks’ balanced salt solution (HBSS) were seeded in a dish with a cover-glass base coated with 0.1% gelatin solution. The energies of the 525 nm and 800 nm excitation light pulses were 2.77 nJ/pulse and 10.58 nJ/pulse, respectively. The sample was scanned over 40 µm with a 1 µm step size along the $x$ and $y$ axes.
Next, to demonstrate the MS-OR-PAM imaging of RBCs and melanoma cells, melanoma-RBC mixed samples were imaged. The RBC suspension was prepared from 10 mL of blood obtained from a Japanese white rabbit (National Defense Medical College Committee for Animal Use approval number 17027). Details of the RBC suspension preparation process were described in a previous publication [24]. The RBC suspension was diluted to 20% with phosphate-buffered saline (PBS). Because the PBS had a high oxygen partial pressure, the hemoglobin in the RBCs was fully oxygenated. Two hours before the PAM imaging, B16-F0 cells in a HBSS were seeded in a dish with a cover-glass base coated with 0.1% gelatin solution containing 0.04% RBCs. The energies of the 525 nm and 800 nm excitation light pulses were 4.00 nJ/pulse and 8.90 nJ/pulse, respectively. The sample was scanned over 100 µm with a 1 µm step size along the $x$ and $y$ axes.
The maximum amplitude projection (MAP) of the three-dimensional PA images was calculated over a thickness of 3000 µm. For spectral subtraction processing, weighted differential images were obtained from the PAM images acquired at 525 nm and 800 nm. A weight of 4.66 was used as this is the ratio between the absorption coefficients of melanosome at 525 nm and 800 nm.
3. RESULTS
A. Optical Simulation of Light Focusing Using the Reflective Objective
The results of the ray-tracing simulation for the propagation of the excitation light through the reflective objective are shown in Fig. 4(a). The collimated excitation light was reflected by the secondary and primary mirrors and then focused on the focal plane. To confirm that the focal length was independent of the wavelength, the dependence of the Strehl ratios on the distance from the objective was calculated at each excitation wavelength. The focal plane ($\Delta z = {0}\;{\unicode{x00B5}{\rm m}}$) was selected by maximizing the Strehl ratio at 525 nm. As shown in Fig. 4(b), the Strehl ratios were maximized at the focal plane at wavelengths of 525 nm and 800 nm, and the values at the focal plane were 1.000 and 0.999, respectively. Thus, we confirmed that the focal length of the reflective objective was independent of wavelength, and no significant aberrations were observed on the focal plane at either 525 nm or 800 nm. Because the incident beam diameter was larger at 800 nm, the Strehl ratio rapidly decreased with defocusing. The profiles of the simulated PSFs are shown in Fig. 4(c), which shows that the defocusing resulted in a reduced spatial resolution and peak power.
Fig. 4. Optical simulation of light focusing using reflective objective. (a) Excitation light (800 nm) ray tracing through the reflective objective. (b) Dependence of Strehl ratios on the distance from the objective. (c) Cross section of point spread function on the planes of $\Delta z = {0}\;\unicode{x00B5}{\rm m}$, 20 µm, and 40 µm at 800 nm. PM, primary mirror; SM, secondary mirror.
Fig. 5. Dependence of FWHM on distance between the objective and microscope calibration scale for (a) a standard objective lens and (b) reflective objective. The values plotted are averages of multiple measurements ($n = {4}$) and error bars indicate standard deviations.
Fig. 6. Multispectral photoacoustic images of melanoma cells. (a) PAM image acquired using the reflective objective at 525 nm. (b) PAM image acquired using the reflective objective at 800 nm. (c) Weighted differential PA image of a melanoma cell acquired using the reflective objective. (d) $Y$ profile of the weighted PA signal intensity acquired using the reflective objective. (e) PAM image acquired using the standard glass objective lens at 525 nm. (f) PAM image acquired using the standard glass objective lens at 800 nm. (g) Weighted differential PA image of a melanoma cell acquired using the standard glass objective. (h) $Y$ profile of the weighted PA signal intensity acquired using the standard glass objective lens. The weighted signal intensity at 800 nm was calculated by multiplication by 4.66, which was the ratio of the absorption coefficients of melanosome in the melanoma cell at 525 nm and 800 nm. A $Y$ profile slice was selected as it contains the region with the largest PA signal at 525 nm.
B. Experimental Evaluation of Focus Shift due to Chromatic Aberration
In Fig. 5(a), the plane with the lowest FWHM differed by 30 µm between 525 nm and 800 nm. From the specification of the standard objective lens that was provided by the manufacturer, the estimated focal shift between 525 nm and 800 nm due to chromatic aberration was 30.21 µm, which coincided with our experimental result. Thus, the focal shift shown in Fig. 5(a) was caused by the chromatic aberration of the standard objective. In contrast, no significant focal shift was observed for the reflective objective, as shown in Fig. 5(b). This result agrees with the simulation results and indicates that the use of a reflective objective enables chromatic aberration-free MS-OR-PAM across the visible-to-NIR wavelength range. Although the focal lengths of the objective lens and incident beam diameter were consistent, the standard objective lens provided better spatial resolution. It is considered that the secondary mirror in the reflective objective caused reduction of the actual numerical aperture, which results in a limited spatial resolution.
C. MS-OR-PAM Imaging of Melanoma Cells and RBCs
The PAM images and weighted differential images of the melanoma cells are shown in Fig. 6. Because the melanosomes have high absorption coefficients at both 525 nm and 800 nm, strong PA signals were detected at both excitation wavelengths. With the reflective objective, similar images were obtained at 525 and 800 nm, as shown in Figs. 6(a) and 6(b), respectively. By contrast, when using the standard glass objective lens, as shown in Figs. 6(e) and 6(f), the change in the optical focal depth resulted in a remarkable decrease in the signal intensity and blurring in the PAM image at 800 nm. This result suggested that the signal decrease and blurring of the PAM image were the result of the focal plane shifting because of chromatic aberration. As shown in Figs. 6(c) and 6(d), when the reflective objective was used to generate images, the melanoma cell signals could be removed by differential processing. However, the signals remained when the standard glass objective lens was used to acquire the images because of the influence of chromatic aberration, as shown in Figs. 6(g) and 6(h).
Fig. 7. Multispectral photoacoustic images of a sample of melanoma cells mixed with RBCs. (a) PAM image acquired using the reflective objective at 525 nm. (b) PAM image acquired using the reflective objective at 800 nm. (c) Superimposed image of oxy-hemoglobin in RBCs and melanosome in melanoma cells acquired by the reflective objective. (d) PAM image acquired using the standard glass objective at 525 nm. (e) PAM image acquired using the standard objective at 800 nm. (f) Superimposed image of ${oxy}$-hemoglobin in RBCs and melanosome in melanoma cells acquired by the standard glass objective.
We measured four individual melanoma cells and statistically evaluated their image pattern and signal intensities. To compare image patterns at 525 nm and 800 nm, the normalized cross correlation (NCC) was calculated. The NCC values decreased from ${0.984}\;{\pm}\;{0.005}$ for the reflective objective to ${0.931}\;{\pm}\;{0.047}$ for the standard glass objective lens. This suggested that the standard glass objective focal shift caused a shift of the image plane. To compare signal intensities at 525 nm and 800 nm, ratios of the signal intensities at the melanoma cells were calculated. The ratio was ${4.02}\;{\pm}\;{1.05}$ for the reflective objective, which was close to the ratio of the absorption coefficients (4.66) of melanosome at 525 nm and 800 nm. The deviation from the absorption coefficient ratio may have originated from the saturation effect of the PA signal intensity [31].
The PAM images of the melanoma cell and RBCs mixture are shown in Fig. 7. As shown in Figs. 7(a) and 7(d), at 525 nm, both the RBCs and the melanoma cell were visualized. As shown in Figs. 7(b) and 7(e), only the melanoma cell was visualized at 800 nm. Because optical absorption coefficient of oxy-hemoglobin decreases 37.8 times from 525 nm to 800 nm, the PA signals from RBC also decrease from 800 to 525 nm. In this situation, as increasing excitation light energy at 800 nm, the PA signals from the melanoma cell increased and reached the saturation level of the measurement system before the PA signals from RBCs became larger than the noise level of the measurement system. Thus, in our experiment, we confirmed that the PA signal produced from RBC was negligibly small at 800 nm and assumed that the PA image acquired at 800 nm visualized only melanoma cells. Figures 7(c) and 7(f) show the superimposed images of the RBCs obtained by weighed differential imaging and melanoma obtained at 800 nm. It is confirmed that there was some cross talk of melanoma signals in the standard glass objective lens. In contrast, the reflective objective could discriminate melanoma from RBCs. In the reflective objective, weighted subtraction increased the ratio of the RBC signal intensities to melanoma signal intensity from 0.46 to 4.66. However, in a standard objective lens, the ratio increased from 0.57 to 1.45.
4. DISCUSSION
As shown in Fig. 4, it is suggested there are no significant chromatic aberration effects when a reflective objective is used in this type of setup. By contrast, the chromatic aberration introduced by standard glass objective lens causes optical focus shifting, which results in a decrease in signal intensity and image blurring, as shown in Fig. 5. In this experiment, the FWHM was enlarged by approximately 4 µm as the optical focus shifted by 30 µm with the use of the standard glass objective lens. If the sample is much larger than the optical focus, the total absorbed energy might be almost unchanged, and thus the change in the PA signal intensity might be limited. However, when samples are smaller or comparable to the optical focus spot size, as is the case for melanosomes in melanoma cells, the total absorbed energy and signal intensity will decrease as the energy density per area decreases. Therefore, the optical focus shifting caused by chromatic aberration can result in significant changes in the intensities of PA signals produced by optical absorbers, as shown in Figs. 6. This problem generates difficulties for spectral processing, such as spectral differential computation [32] and spectral fitting [24], which compare the PA signal intensities measured at multiple wavelengths to obtain information related to the optical absorption spectrum of the absorber. Thus, in MS-OR-PAM using a standard objective lens, chromatic aberration effects should be avoided, for example, by limiting the wavelength range so that the influence of chromatic aberration is negligible, by performing deblurring processing on acquired images [33,34], by using optics that compensate for the effect of chromatic aberration, or by adjusting the objective lens position at each wavelength. For correction processing after image acquisition, most techniques have been developed for planar image acquisition techniques, and technical improvement is required for volumetric imaging techniques, including OR-PAM. Therefore, at present, it is difficult to completely remove the effects of chromatic aberration via correction processing. Compensation for the effects of chromatic aberration is possible using adaptive optics such as variable-focus lenses or physical adjustment of the objective lens position, but to the best of our knowledge, the response time of such techniques is longer than the repetition period of the SC light source. In MS-OR-PAM, it is necessary to minimize temporal and spatial deviation to make accurate comparisons between wavelengths. Although fast MS-OR-PAM techniques have been studied for this purpose [35], it is difficult to incorporate these techniques to achieve fast MS-OR-PAM. However, when a reflective objective is used, the problem of chromatic aberration can be physically avoided. Thus, this technique has a considerable advantage as it avoids the necessity of taking any of these measures to compensate for chromatic aberration. This approach should prove especially advantageous in the future for applications such as spectral unmixing with MS-OR-PAM imaging.
Most of the PAM systems reported so far have used only narrow wavelength ranges, in the visible or NIR, for imaging. By using our system to perform MS-OR-PAM over a wide spectral bandwidth from the visible to the NIR to generate images that do not suffer from the effects of chromatic aberration, it should prove possible to target a wide range of endogenous optical absorbers, including blood, lipids, and collagen, as well as various exogenous contrast agents such as indocyanine green.
5. CONCLUSION
In this study, we developed and evaluated the performance of an MS-OR-PAM system designed to enable micro-scale observation over a wide wavelength range without the influence of chromatic aberration. Using the MS-OR-PAM device, we were able to acquire micro-scale PA biological tissue images over a wide wavelength range. The system can be used to carry out spectral unmixing and may be applicable to the discrimination of abnormal tissue from normal body tissue by analysis of different components within the living body on the cell-to-tissue scale.
Funding
Princess Takamatsu Cancer Research Fund (15-24703); Ministry of Education, Culture, Sports, Science and Technology (MEXT) (21H00445, 19H05436); Japan Society for the Promotion of Science (JSPS) (19K12856); Ministry of Defense- Japan (Defense Medicine Basic Research Program (A), Defense Medicine Basic Research Program (C)).
Acknowledgment
The authors thank H. Sanguu (Yokogawa Electric Corporation) for his important advice regarding this research and for lending us parts of the experimental setup. The authors also thank A. Shito, Y. Inoue, and Y. Tabata (SigmaKoki Co., Ltd) for optical design and simulation assistance. The authors also thank K. Tsujita for his important contribution to this work, T. Kushibiki for lending us a photoacoustic sensor, and M. Miyashita for preparation of the melanoma cell sample. The authors also thank J. Ukon (UKON Craft Science Ltd.) for supporting the light source operation. The research was supported by Grants-in-Aid for Scientific Research on Innovative Areas “Singularity Biology (No.8007)” from MEXT awarded to MI; JSPS KAKENHI awarded to TH, SO, and MI; Defense Medicine Basic Research Program (A) awarded to MI, SO, and TH; Defense Medicine Basic Research Program (C) awarded to TH; Princess Takamatsu Cancer Research Fund awarded to MI.
Disclosures
MI: Yokogawa Electric Corporation (F).
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.
REFERENCES
1. W. Song, Q. Wei, W. Liu, T. Liu, J. Yi, N. Sheibani, A. A. Fawzi, R. A. Linsenmeier, S. Jiao, and H. F. Zhang, “A combined method to quantify the retinal metabolic rate of oxygen using photoacoustic ophthalmoscopy and optical coherence tomography,” Sci. Rep. 4, 6525 (2014). [CrossRef]
2. N. J. M. Haven, P. Kedarisetti, B. S. Restall, and R. J. Zemp, “Reflective objective-based ultraviolet photoacoustic remote sensing virtual histopathology,” Opt. Lett. 45, 535–538 (2020). [CrossRef]
3. Y. He, L. Wang, J. Shi, J. Yao, L. Li, R. Zhang, C. H. Huang, J. Zou, and L. V. Wang, “In vivo label-free photoacoustic flow cytography and on-the-spot laser killing of single circulating melanoma cells,” Sci. Rep. 6, 39616 (2016). [CrossRef]
4. Y. Wang, K. Maslov, Y. Zhang, S. Hu, L. Yang, Y. Xia, J. Liu, and L. V. Wang, “Fiber-laser-based photoacoustic microscopy and melanoma cell detection,” J. Biomed. Opt. 16, 011014 (2011). [CrossRef]
5. M. K. Dasa, C. Markos, M. Maria, C. R. Petersen, P. M. Moselund, and O. Bang, “High-pulse energy supercontinuum laser for high-resolution spectroscopic photoacoustic imaging of lipids in the 1650–1850 nm region,” Biomed. Opt. Express 9, 1762–1770 (2018). [CrossRef]
6. J. Yao, K. I. Maslov, Y. Zhang, Y. Xia, and L. V. Wang, “Label-free oxygen-metabolic photoacoustic microscopy in vivo,” J. Biomed. Opt. 16, 076003 (2011). [CrossRef]
7. L. Wang, K. Maslov, and L. V. Wang, “Single-cell label-free photoacoustic flowoxigraphy in vivo,” Proc. Natl. Acad. Sci. USA 110, 5759–5764 (2013). [CrossRef]
8. E. I. Galanzha, Y. A. Menyaev, A. C. Yadem, M. Sarimollaoglu, M. A. Juratli, D. A. Nedosekin, S. R. Foster, A. Jamshidi-Parsian, E. R. Siegel, I. Makhoul, L. F. Hutchins, J. Y. Suen, and V. P. Zharov, “In vivo liquid biopsy using Cytophone platform for photoacoustic detection of circulating tumor cells in patients with melanoma,” Sci. Transl. Med. 11, eaat5857 (2019). [CrossRef]
9. E. Auksorius, B. R. Boruah, C. Dunsby, P. M. Lanigan, G. Kennedy, M. A. Neil, and P. M. French, “Stimulated emission depletion microscopy with a supercontinuum source and fluorescence lifetime imaging,” Opt. Lett. 33, 113–115 (2008). [CrossRef]
10. H. Kano, “Molecular vibrational imaging of a human cell by multiplex coherent anti-Stokes Raman scattering microspectroscopy using a supercontinuum light source,” J. Raman Spectrosc. 39, 1649–1652 (2008). [CrossRef]
11. C. Lefort, R. P. O’Connor, V. Blanquet, L. Magnol, H. Kano, V. Tombelaine, P. Lévêque, V. Couderc, and P. Leproux, “Multicolor multiphoton microscopy based on a nanosecond supercontinuum laser source,” J. Biophoton. 9, 709–714 (2016). [CrossRef]
12. C. Lee, S. Han, S. Kim, M. Jeon, M. Y. Jeon, C. Kim, and J. Kim, “Combined photoacoustic and optical coherence tomography using a single near-infrared supercontinuum laser source,” Appl. Opt. 52, 1824–1828 (2013). [CrossRef]
13. M. Bondu, C. Brooks, C. Jakobsen, K. Oakes, P. M. Moselund, L. Leick, O. Bang, and A. Podoleanu, “High energy supercontinuum sources using tapered photonic crystal fibers for multispectral photoacoustic microscopy,” J. Biomed. Opt. 21, 061005 (2016). [CrossRef]
14. X. Shu, M. Bondu, B. Dong, A. Podoleanu, L. Leick, and H. F. Zhang, “Single all-fiber-based nanosecond-pulsed supercontinuum source for multispectral photoacoustic microscopy and optical coherence tomography,” Opt. Lett. 41, 2743–2746 (2016). [CrossRef]
15. M. Bondu, M. Denninger, P. M. Moselund, and A. Podoleanu, “Using a single supercontinuum source for visible multispectral photoacoustic microscopy and 1300 nm optical coherence tomography,” Proc. SPIE 10415, 1041507 (2017). [CrossRef]
16. M. Bondu, M. J. Marques, P. M. Moselund, G. Lall, A. Bradu, and A. Podoleanu, “Multispectral photoacoustic microscopy and optical coherence tomography using a single supercontinuum source,” Photoacoustics 9, 21–30 (2018). [CrossRef]
17. M. K. Dasa, G. Nteroli, P. Bowen, G. Messa, Y. Feng, C. R. Petersen, S. Koutsikou, M. Bondu, P. M. Moselund, A. Podoleanu, A. Bradu, C. Markos, and O. Bang, “All-fibre supercontinuum laser for in vivo multispectral photoacoustic microscopy of lipids in the extended near-infrared region,” Photoacoustics 18, 100163 (2020). [CrossRef]
18. B. Rao, F. Soto, D. Kerschensteiner, and L. V. Wang, “Integrated photoacoustic, confocal, and two-photon microscope,” J. Biomed. Opt. 19, 036002 (2014). [CrossRef]
19. R. Cao, J. P. Kilroy, B. Ning, T. Wang, J. A. Hossack, and S. Hu, “Multispectral photoacoustic microscopy based on an optical–acoustic objective,” Photoacoustics 3, 55–59 (2015). [CrossRef]
20. C. Zhang, Y. S. Zhang, D. K. Yao, Y. Xia, and L. V. Wang, “Label-free photoacoustic microscopy of cytochromes,” J. Biomed. Opt. 18, 020504 (2013). [CrossRef]
21. P. Hajireza, A. Forbrich, and R. Zemp, “In-vivo functional optical-resolution photoacoustic microscopy with stimulated Raman scattering fiber-laser source,” Biomed. Opt. Express 5, 539–546 (2014). [CrossRef]
22. M. M. Kabir, A. M. Choubal, and K. C. Toussaint Jr., “Application of a reflective microscope objective for multiphoton microscopy,” J. Microsc. 271, 129–135 (2018). [CrossRef]
23. H. Wang, X. Yang, Y. Liu, B. Jiang, and Q. Luo, “Reflection-mode optical-resolution photoacoustic microscopy based on a reflective objective,” Opt. Express 21, 24210–24218 (2013). [CrossRef]
24. T. Hirasawa, R. J. Iwatate, M. Kamiya, S. Okawa, Y. Urano, and M. Ishihara, “Multispectral photoacoustic imaging of tumours in mice injected with an enzyme-activatable photoacoustic probe,” J. Opt. 19, 014002 (2017). [CrossRef]
25. K. M. Artyukov and K. Krymski, “Schwarzschild objective for soft x-rays,” Opt. Eng. 39, 2163–2170 (2000). [CrossRef]
26. I. A. Artyukov, “Schwarzschild objective and similar two-mirror systems,” Proc. SPIE 8678, 86780A (2012). [CrossRef]
27. G. Lan and M. D. Twa, “Theory and design of Schwarzschild scan objective for optical coherence tomography,” Opt. Express 27, 5048–5064 (2019). [CrossRef]
28. S. A. Prahl, “Optical absorption of hemoglobin,” 1999, http://omlc.org/spectra/hemoglobin/index.html.
29. S. L. Jacques, “Optical properties of biological tissues: a review,” Phys. Med. Biol. 58, R37–R61 (2013). [CrossRef]
30. K. Jimbow, Y. Miyake, K. Homma, K. Yasuda, Y. Izumi, A. Tsutsumi, and S. Ito, “Characterization of melanogenesis and morphogenesis of melanosomes by physicochemical properties of melanin and melanosomes in malignant melanoma,” Cancer Res. 44, 1128–1134 (1984).
31. J. Yao and L. V. Wang, “Sensitivity of photoacoustic microscopy,” Photoacoustics 2, 87–101 (2014). [CrossRef]
32. T. Hirasawa, R. J. Iwatate, M. Kamiya, S. Okawa, M. Fujita, Y. Urano, and M. Ishihara, “Spectral-differential-based unmixing for multispectral photoacoustic imaging,” Appl. Opt. 57, 2383–2393 (2018). [CrossRef]
33. S. J. Chen and H. L. Shen, “Multispectral image out-of-focus deblurring using interchannel correlation,” IEEE Trans. Image Process. 24, 4433–4445 (2015). [CrossRef]
34. Z. Sadeghipoor, Y. M. Lu, and S. Süsstrunk, “Gradient-based correction of chromatic aberration in the joint acquisition of color and near-infrared images,” Proc. SPIE 9404, 94040F (2015). [CrossRef]
35. J. Chen, Y. Zhang, X. Li, J. Zhu, D. Li, S. Li, C.-S. Lee, and L. Wang, “Confocal visible/NIR photoacoustic microscopy of tumors with structural, functional, and nanoprobe contrasts,” Photon. Res. 8, 1875–1880 (2020). [CrossRef]
