Open Access
Add to BibSonomy
Abstract
Optical-resolution photoacoustic microscopy (OR-PAM) has been widely used for imaging blood vessel and oxygen saturation of hemoglobin (sO2), providing high-resolution functional images of living animals in vivo. However, most of them require one or multiple bulky and costly pulsed lasers, hindering their applicability in preclinical and clinical settings. In this paper, we demonstrate a reflection-mode low-cost high-resolution OR-PAM system by using two cost-effective and compact laser diodes (LDs), achieving microvasculature and sO2 imaging with a high lateral resolution of ∼6 µm. The cost of the excitation sources has dramatically reduced by ∼20–40 times compared to that of the pulsed lasers used in state-of-the-art OR-PAM systems. A blood phantom study was performed to show a determination coefficient R2 of 0.96 in linear regression analysis. Experimental results of in vivo mouse ear imaging show that the proposed dual-wavelength LD-based PAM system can provide high-resolution functional images at a low cost.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Optical-resolution photoacoustic microscopy (OR-PAM) is a label-free and noninvasive imaging technique that converts absorbed photon energy into mechanical energy (PA signals), providing sub-cellular-resolution images with optical absorption contrast [1,2]. For blood vessel imaging in OR-PAM, two dominant intrinsic absorbers, oxy- and deoxy-hemoglobin, absorb the photon energy of light pulses and then generate PA signals through thermoelastic expansion. The PA signals provide functional information of living tissue, such as oxygen saturation (sO2) and metabolic rate of oxygen [3–11], which are essential parameters for disease diagnosis [12]. For example, angiogenesis with high sO2 usually happens in the vicinity of tumors, while in core regions, the abnormal microvasculature has overall low oxygenation because of the tortuous blood vessels and interrupted blood flow [13]. Therefore, although OR-PAM has limited imaging depth (<1 mm), it has been widely applied to cancer screening, diagnosis, and prognosis, especially for superficial tissues, such as the skin and gastrointestinal tract [6,8,14]. However, since a short laser pulse is required to induce transient local pressure rise in PAM, a costly and bulky laser, such as Q-switched laser diode-pumped solid-state laser, is commonly needed, which hinders the wide usage of OR-PAM in preclinical and clinical applications. In this regard, compact and low-cost laser diodes (LDs) have been investigated as alternative excitation sources in OR-PAM. For example, a pulsed LD at 905 nm with a pulse energy of 3 µJ and a repetition rate of 1 kHz has been demonstrated as an excitation source for in-vivo mouse ear imaging [15]. However, the high lateral resolution of 7 µm was achieved by focusing the laser beam using a 60${\times} $ objective lens with a high numerical aperture (NA) of 0.7, resulting in a short working distance and implemented in a transmission mode. Moreover, because of the low signal-to-noise ratio (SNR), 128 signal averaging was required to improve the image quality, which would slow down the imaging speed. A laser scanning PAM system with pulsed LD at 905 nm and pulse energy of 13 µJ has also been demonstrated for ex-vivo mouse ear imaging without the requirement of signal averaging, thus, improving the imaging speed [16]. However, an aspheric lens with a high NA of 0.71 is needed to focus the LD beam on samples to achieve a lateral resolution of 21 µm, which limits the working distance. Besides, the lateral resolution is not high enough to resolve microvascular networks. In addition to using pulsed LDs, continuous-wave (CW) LDs operated at pulsed mode have also been investigated in PA imaging. For instance, a CW LD at 450 nm was pulsed driven to achieve a high repetition rate of 625 kHz and pulse energy of ∼200 nJ, serving as an excitation source for in-vivo PAM imaging of mouse ear [17]. Nevertheless, the system requires 500 signal averaging, resulting in only 1.25k A-lines per second. Besides, the resolution of the proposed system is ∼110 µm, which should be further improved to image microvasculature. Although phantom studies and animal experiments have been carried out to demonstrate that LDs can be an option in PAM imaging, most of the existing LD-based PAM systems suffer from low spatial resolution and low imaging speed [15–20], and thus they have limited practical applications. Recently, we developed a high-resolution (∼4.8 µm) and high-speed (30k Hz A-line rate) LD-based PAM system by optimizing the optical system design for LD beam reshaping with the consideration of the LD emitter size, achieving high-quality in-vivo microvasculature imaging [21]. However, only a single wavelength LD was used in the previous system, and hence, the essential diagnostic parameter (e.g., sO2) cannot be estimated. Moreover, since a high NA objective lens with a short working distance of 4.1 mm is used to achieve the high resolution, the system can only be implemented in a transmission mode, which is only suitable for imaging thin tissues. A spectroscopic PA sensor with dual-wavelength LDs has recently been developed for local vascular sO2 measurement, but it can only be operated at single-point detection with a low resolution of ∼110 µm [22], and thus, the sO2 values between neighboring arteries and veins would be mixed. Therefore, the performance and applications of the LD-based spectroscopic PA sensor are limited. To the best of our knowledge, there is no previous study showing that an LD-based PAM system can achieve high-resolution sO2 imaging for living tissue.
To overcome the aforementioned challenges, in this paper, we demonstrate a reflection-mode low-cost high-resolution OR-PAM system for in-vivo sO2 imaging using two laser diodes with different wavelengths, which are operating at a pulsed mode. As LDs are not perfect point sources, the emitter sizes and light beam shapes of the LDs are taken into consideration when optimizing the optical system design to reshape the LD beams, hence, achieving a high lateral resolution of ∼6 µm. Blood phantom study was first performed to evaluate the quantification ability of our system for sO2 measurement, and then in-vivo sO2 imaging of mouse ear was conducted to validate the performance of the proposed low-cost dual-wavelength LD-based PAM.
2. Method
2.1 Dual-wavelength pulsed LD sources
To quantify sO2 via a spectral unmixing algorithm, at least two laser pulses with different wavelengths are required [13]. In this paper, two CW LDs, a blue laser diode (NDB7675, Opt Laser Inc.) and a green laser diode (FVLD-520-1000M, Frankfurt Laser Inc.), are each pulsed driven by a pulse driver (LMG1020EVM, Texas Instruments Inc.) with a repetition rate of 10 kHz, acting as excitation sources in OR-PAM. The output pulse widths and spectra of the two LDs are shown in Fig. 1. From Fig. 1(a), we can see that the output pulses of the two LDs have the same pulse width of ∼6.8 ns, which are detected by a photodiode (PDA10A2, Thorlabs Inc.) and read out by an oscilloscope (DS1102E, RIGOL Technologies Inc.). The time delay for the two output LD pulses is 500 ns, which can be easily preset or adjusted by the delay time of triggering signals for the two pulse drivers. The output spectra of the two LDs are measured by a spectrometer (SP-RES-UV-Vis-NIR, Sarspec, Lda), showing that central wavelengths of the blue and green LDs are 458 nm and 517 nm, respectively (Fig. 1(b)). The molar extinction coefficients of the oxyhemoglobin (HbO2) and deoxyhemoglobin (HbR) against wavelengths are shown as the red and light blue dotted lines in Fig. 1(b), respectively. The values corresponding to the central wavelengths of the LDs output would be used to quantify sO2 via the spectral unmixing algorithm [13]. To evaluate the pulse-to-pulse stability, the pulse energies of the two LDs were measured during two hours of continuous operation in the pulsed mode. The standard deviation (SD) of the pulse energy of the blue and green LDs are ∼3.4% and ∼2.5%, respectively, indicating the high stability of the LDs. The pulse energies of the blue and green LDs are ∼32 nJ and ∼34 nJ, respectively, for both in-vitro and in-vivo experiments. The optical fluences on the tissue surface are estimated to be ∼1.6 mJ/cm2 and ∼1.7 mJ/cm2 (assuming the optical focus is ∼200 µm beneath the skin surface), respectively, which are both lower than that of the safety limit (20 mJ/cm2) suggested by the American National Standards Institute [23].
Fig. 1. The output pulse widths and spectra of the blue and green LDs operated in a pulsed mode. (a) The output pulse profiles of the two LDs. (b) The output spectra of the two LDs and the molar extinction coefficients of the HbO2 and HbR against wavelengths. LD: laser diode; HbO2: oxyhemoglobin; HbR: deoxyhemoglobin.
2.2 Dual-wavelength LD-based PAM system
In OR-PAM imaging, the high lateral resolution is achieved by tightly focusing a light beam on the samples, which is typically generated from a laser source with high beam quality (${\textrm{M}^2} \approx 1$, a Gaussian TEM00 spatial mode). However, since the two LDs have an emitter size of 15 µm ${\times} $ 1 µm (vertical ${\times} $ horizontal), they are not ideal point sources for high-resolution imaging, and thus a diffraction-limited focal spot is hard to achieve. To obtain high-resolution LD-based PA images, the emitter size of the LDs should be taken into consideration when reshaping the LD beam. In our previous study, we have successfully demagnified the vertical dimension of an LD emitter by optimizing the optical system design, achieving a high lateral resolution of ∼4.8 µm [21]. In brief, as shown in Fig. 2(a), lens 1 (L1) is used to collimate the LD beam and lens 4 (L4) is an objective lens that is used to focus the LD beam on the samples. Between these two lenses, two cylindrical lenses (L2 and L3) are used to expand the LD beam in the vertical dimension. Using geometrical optics, the focal spot size of the LD on the samples (placed on the image plane) can be expressed as: $h^{\prime} = ({{f_4}/{f_3}} )\cdot ({{f_2}/{f_1}} )\cdot h$, where ${f_i}$ is the focal length of the lens ${\textrm{L}_i}$ ($i\; $= 1, 2, 3, 4), and h is the size of the LD emitter. Therefore, by determining the focal lengths of the four lenses, a small focal spot can be generated on the samples, achieving high-resolution LD-based PAM imaging.
Fig. 2. The optical system design of low-cost high-resolution LD-based PAM. (a) The diagram of the optical system that demagnifies the LD emitter to achieve high-resolution imaging. (b) The system schematic of the dual-wavelength LD-based PAM. (c, i)–(c, iii) The beam shapes of the green LD after being reshaped by AL, CL2, and the iris, respectively. L: lens; IP: image plane; $f$: focal length; AL: aspheric lens; CL: cylindrical lens.
To build upon our previous research study, here, we develop a low-cost high-resolution dual-wavelength LD-based PAM system, shown in Fig. 2(b). The laser beam from the green LD is first collimated by an aspheric lens (A240TM-A, ${f_1}$ = 8 mm, Thorlabs, Inc.), obtaining an elliptical beam (shown in Fig. 2(c, i)) due to the different divergence angles on vertical and horizontal directions (corresponding to the $y$- and $z$-axis in Fig. 2(b)). Then, two cylindrical lenses (GLH15-025-025, ${f_2}$ = 25 mm, Hengyangguangxue; LJ1267RM-A, ${f_3}$ = 250 mm, Thorlabs, Inc.) are applied to expand the LD beam on the $z$-axis, generating a rectangular beam (shown in Fig. 2(c, ii)). Subsequently, an iris (∼7 mm in diameter) is placed after the second cylindrical lens to reshape the light to be a circular beam (shown in Fig. 2(c, iii)). Since the blue LD has the same emitter size as the green one, the blue LD beam is reshaped with the same optical setup with the same parameters. The green light and the blue light are then combined through a dichroic mirror (DMSP490, Thorlabs Inc.), and finally focused by an objective lens (LMU-5X-NUV, ${f_4}$ = 39.9 mm, Thorlabs Inc.) on a sample after passing through the empty hole of a ring-shaped ultrasonic transducer (UT) (Focal length: 6.3 mm; center frequency: 40 MHz; -6 dB bandwidth: 84%). The long working distance (37.5 mm) of the objective lens and the ring-shaper UT allows the system to be implemented in a reflection mode, enabling thick tissue in-vivo imaging. The sample is placed on a sample holder with a transparent window which is covered by a thin membrane to prevent water leakage from the water tank to the sample holder. The excited PA signals propagate from the sample to the ring-shaped UT. Two motorized stages (L-509.10SD00, PI (Physik Instrumente Singapore LLP) are used to translate the sample for raster scanning.
As the LDs are not ideal point sources, the blue and green LD beams after the second cylindrical lenses (CL2 in Fig. 2(b)) would not be perfect collimated beams and have different divergence angles. Therefore, the focal planes of the two LD beams after being focused by the objective lens cannot be guaranteed to be the same on the sample, resulting in the PA signals excited by the two LDs being generated from different depths, hence, affecting the accuracy of the sO2 measurement. The achromatic objective lens could also introduce an axial focal shift of ∼24 µm due to the wavelength differences of the two LDs. To ensure the two LD beams are focused on the same focal plane, two identical lenses (LA1433-A, Thorlabs Inc.) are inserted after the second cylindrical lens on the blue LD light path (shown in Fig. 2(b), the red dotted region). The two lenses are first aligned as a 4-f system (a relay lens), and then the position of lens 4 is tuned by a translation stage (MT1B, Thorlabs Inc.) to change the focal plane of the blue LD focal spot to meet that of the green LD focal spot. From the focal lengths of the aspheric lens, cylindrical lenses 2 and 3, the objective lens, and the LD emitter size, the theoretical resolution of the proposed system can be estimated as $h^{\prime} = ({39.9/250} )\cdot ({25/8} )\cdot 15\; {\mathrm{\mu} \mathrm{m}} \approx 7.5\mathrm{\;\ \mu m}$.
2.3 sO2 calculation from PA signals
The sO2 is defined as the percentage of the molar concentration of HbO2 over the molar concentration of total hemoglobin, which can be expressed as [13]:
where the ${C_{Hb{O_2}}}$ and ${C_{HbR}}$ are the molar concentrations of HbO2 and HbR, respectively. When two wavelengths of laser pulses are applied to measure the sO2, a linear spectral fitting method (e.g., the least square method) is commonly used, and the PA amplitudes of the PA signals can be simplified as [13,24,25]:Here, the $\varepsilon _{{\lambda _i}}^{Hb{O_2}}$ and $\varepsilon _{{\lambda _i}}^{HbR}$ are the known molar extinction of HbO2 and HbR at specific wavelengths ($i = 1, 2$), respectively, which can be read from Fig. 1(b). ${\mathrm{\Phi }_{{\lambda _i}}}$ is the local optical fluence. Using laser pulses with two different wavelengths, the ${C_{Hb{O_2}}}$ and ${C_{HbR}}$ can be solved as:
Finally, by substituting Eq. (3) into Eq. (1), we can estimate the sO2.
3. Results
3.1 Resolution measurement
To measure the lateral resolution of our LD-based PAM system, a sharp edge was imaged, as shown in Fig. 3. Since the LD emitters are rectangular, the resolution is evaluated on both the x- and y-axis, corresponding to the vertical and horizontal dimensions of the LD emitters being focused on samples. We first fit the measured data (blue dots in Fig. 3(a)–(d)) with an error function to obtain an edge spread function (ESF) (blue dotted lines in Fig. 3(a)–(d)), then the first-order derivative of the ESF was calculated to obtain the line spread function (LSF) (red lines in Fig. 3(a)–(d)). The full width at half maximum (FWHM) of the LSF is measured to evaluate the lateral resolution. From Fig. 3(a) and (c), we can see that for the blue LD beam, the lateral resolutions on the x- and y-axis are ∼6.0 µm and ∼5.6 µm, respectively. Whereas for the green LD beam, the lateral resolutions on the x- and y-axis are ∼6.1 µm and ∼6.2 µm, respectively (Fig. 3(b) and (d)). Therefore, our dual-wavelength LD-based PAM system has a high lateral resolution of ∼6 µm. The measured lateral resolution is slightly better than our theoretical value (7.5 µm) likely because the irises placed behind the second cylindrical lens serve as a vignetting stop which blocks the peripheral light emitted from the edge of the LD emitter at the expense of optical energy, and thus, reducing the actual size of the LD emitter being focused.
Fig. 3. Resolution measurement of the dual-wavelength LD-based PAM system. (a), (b) The lateral resolution along the x-axis measured with the blue and green LDs, respectively. (c), (d) The lateral resolution along the y-axis measured with the blue and green LDs, respectively. The left insets of (a)–(d) show the LD-based PAM image of the sharp edge. (e), (f) The axial resolution measured with the blue and green LDs, respectively.
The enveloped A-line signal is extracted to evaluate the axial resolution of our system, as shown in Fig. 3(e) and (f). The FWHM of the enveloped PA signals excited by both the blue and green LD beams are almost the same, which is ∼40 µm.
3.2 In-vitro evaluation of the dual-wavelength LD-based PAM system
To evaluate the ability of the proposed dual-wavelength LD-based PAM system for sO2 measurement, blood phantoms were tested in vitro. Fresh mouse blood was collected and mixed with an anticoagulant. Five mouse blood samples with different sO2 values (0, 0.25, 0.5, 0.75, and 1) were prepared by adding different dosages of sodium dithionite, which can change the sO2 values from 1 to 0 with a concentration of 2.5 mg/ml (sodium dithionite/blood) [26]. 30 min after adding the sodium dithionite, each of the blood samples was then poured into a container that has a transparent thin membrane at the bottom to allow LD beam and PA wave propagation, and was sealed by a plastic cover on the top. The air in the container was replaced by nitrogen gas. Finally, the container with blood samples was placed on the sample holder and imaged by our system. Since the oxygenation level of the blood samples is stable for a short time (within 15 min) [26], the whole process of imaging blood samples was performed within 2 min to obtain the optimal results. The sO2 results of the blood phantoms measured by our system are shown in Fig. 4. Each measured sO2 value was averaged $6 \times {10^4}$ times from different positions on the sample. With the repetition rate of 10 kHz, the time for data collection for each phantom is ∼6 s. As shown in Fig. 4, the measured values have a strong linear relationship with the preset ones with a determination coefficient ${R^2}$ of 0.96, which is comparable to the conventional PAM systems [27,28].
3.3 In-vivo mouse ear imaging
To evaluate the performance of the proposed low-cost dual-wavelength LD-based PAM system for high-resolution sO2 imaging, a mouse ear was imaged. The animal experiment protocol was approved by the Animal Ethics Committee at The Hong Kong University of Science and Technology. The mouse was maintained anesthetized with ∼1.5% isoflurane mixed with oxygen at a flow rate of 0.6 L/min during the experiments. The scanning step size along the x-axis is 0.94 µm, which is limited by the velocity of the motorized stage. The scanning step size along the y-axis is 3.1 µm, which is about half of the resolution. The LD-based PAM images of a mouse ear under the blue and green LD excitations are shown in Fig. 5(a) and (b), respectively, where the microvasculature networks can be clearly observed. The SNRs of our LD-based PAM system can reach 31.6 dB and 28.3 dB for the 458 nm and 517 nm LDs, respectively. As shown in Fig. 1(b), the molar extinction coefficient of the HbO2 at 458 nm is ∼2 times higher than that of 517 nm, while the molar extinction coefficient of the HbR at 458 nm is ∼0.87 times of that at 517 nm. Therefore, the PAM images acquired using 458 nm (Fig, 5(a)) are brighter than that of 517 nm (Fig. 5(b)), especially on the arteries. From the depth-encoded images as shown in Fig. 5(c) and (d), we can see that the PA images of microvascular networks under the blue and green LDs show similar depth information, indicating that the two LD beams are focused on the same plane. Then, the values on each pixel of a high-resolution sO2 image can be calculated by using the Eqs. (1)–(4). To eliminate the effect of the random values on the background and enhance the visualization of blood vessels in the calculated sO2 image, the 458 nm PAM image (Fig. 5(a)) is first processed by using morphological operation and top-hat transform to enhance the blood vessel and suppress the background information [29,30]. Then, a threshold is applied to the enhanced image to obtain a binary mask, which is eventually employed in the calculated sO2 image to obtain a background-free image, as shown in Fig. 5(e). The average sO2 of the arteries and veins of the three marked white lines (R1–R3 in Fig. 5(e)) are shown in Fig. 5(f), which is consistent with the typical oxygen saturation of arteries (over 95%) and veins (between 65% and 75%) under normal conditions [31].
Fig. 5. In-vivo dual-wavelength high-resolution LD-based PAM imaging of a mouse ear. (a) and (b) LD-based PAM images of a mouse ear with blue and green LD excitations, respectively. (c) and (d) Depth-encoded images of the mouse ear with blue and green LD excitations, respectively. (e) sO2 image of the mouse ear. (f) Average sO2 of the arteries (orange bars) and veins (blue bars) of the three marked white lines (R1–R3) in (e). Error bars are the SDs.
4. Discussions
In this paper, we demonstrate a reflection-mode low-cost high-resolution dual-wavelength LD-based PAM system by over-driving two CW LDs, achieving high-quality microvasculature and sO2 imaging. Compared with most of the existing LD-based PAM systems, the high resolution of ∼6 µm of our system is achieved by well designing and optimizing the optical system to demagnify the emitter size of the LDs. The long working distance of the low NA objective lens and the ring-shaped UT allow the reflection-mode LD-based PAM system to be implemented, enabling in-vivo thick tissue imaging. While in other previously demonstrated high-resolution LD-based PAM systems, most of them are implemented in a transmission mode, and a high NA objective lens with limited working distance is usually needed, preventing them from imaging thick tissue [15,16,21]. The price of the blue and green LDs is USD 137.00 and USD 460.00, respectively. Each pulsed driver used in this paper costs USD 198.00. Thus, the total price of our excitation source system is USD 993.00, which is ∼20–40 times cheaper than the laser that is commonly used in conventional OR-PAM (USD 20K–40 K in general). Besides, with the high compactness of the LDs (11 mm in length and 9 mm in diameter) and the pulse driver (49 mm (length) ${\times} $ 40 mm (width)${\times} $ 22 mm (height)), our excitation source system can be packaged in small size. Although our LD-based PAM system is currently occupying a relatively large space, mainly because of the long focal lengths of the cylindrical lenses and the two identical lenses in the blue LD light path, a more compact optical system can be achieved by using short focal lengths of lenses. Thus, a portable LD-based PAM system can be potentially developed. Moreover, our LD-based system is found to be stable in long term, which has also been demonstrated in Ref. [17]. With the dramatic price reduction, more compact in size, high stability, and superior imaging performance, our LD-based PAM system has a high potential to be commercialized, promoting the wide usage of PAM in clinical and preclinical applications.
Although our current system has shown promising high-resolution sO2 imaging, the performance of the system can be further improved. First, the imaging speed can be further increased. Currently, the repetition rate of the two LDs is 10 kHz. However, they can be operated at 50 kHz without being damaged and degraded. Therefore, large field-of-view LD-based PAM imaging with high imaging speed can be achieved by using fast scanning strategies, such as a microelectromechanical system [32], a hexagon-mirror scanner [4,7], and a two-axis torsion-bending scanner [33]. Second, the compactness of the system can be further improved as the current system requires a set of lenses to reshape the LD beams. To achieve a portable imaging probe, optical fiber can be adopted to deliver the LD beams in future work. Third, LDs with more wavelengths can be incorporated to broaden the applications of LD-based PAM, revealing different microstructures in tissues.
5. Conclusions
In summary, we have developed a reflection-mode low-cost high-resolution dual-wavelength LD-based PAM system for accurate sO2 imaging, which, to the best of our knowledge, is the first demonstration of high-resolution sO2 imaging using LDs. Two CW LDs (blue and green) are pulsed driven as excitation sources, and then the emitter sizes of LDs are taken into account to optimize the shape of the LD beams, achieving a high resolution of ∼6 µm. The blood phantom study shows that our system has a strong linear correlation (${R^2} = 0.96$) with the preset sO2 values. The in-vivo mouse ear imaging demonstrated that our system could provide high-quality microvasculature and sO2 imaging. With the significant cost reduction, superior performance in functional imaging, and potentially high compactness, the proposed dual-wavelength LD-based PAM is expected to promote the wide application of PAM in preclinical and clinical applications.
Funding
Innovation and Technology Commission (ITS/036/19).
Disclosures
1. V. T. C. T. and T. T. W. W. have a financial interest in PhoMedics Limited, which, however, did not support this work. 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. J. Yao and L. V. Wang, “Photoacoustic microscopy,” Laser & Photonics Reviews 7(5), 758–778 (2013). [CrossRef]
2. L. V. Wang and S. Hu, “Photoacoustic tomography: in vivo imaging from organelles to organs,” Science 335(6075), 1458–1462 (2012). [CrossRef]
3. J. Yao, L. Wang, J.-M. Yang, K. I. Maslov, T. T. W. Wong, L. Li, C.-H. Huang, J. Zou, and L. V. Wang, “High-speed label-free functional photoacoustic microscopy of mouse brain in action,” Nat. Methods 12(5), 407–410 (2015). [CrossRef]
4. B. Lan, W. Liu, Y. Wang, J. Shi, Y. Li, S. Xu, H. Sheng, Q. Zhou, J. Zou, U. Hoffmann, W. Yang, and J. Yao, “High-speed widefield photoacoustic microscopy of small-animal hemodynamics,” Biomed. Opt. Express 9(10), 4689 (2018). [CrossRef]
5. C. Liu, Y. Liang, and L. Wang, “Optical-resolution photoacoustic microscopy of oxygen saturation with nonlinear compensation,” Biomed. Opt. Express 10(6), 3061 (2019). [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(7), 076003 (2011). [CrossRef]
7. J. Chen, Y. Zhang, L. He, Y. Liang, and L. Wang, “Wide-field polygon-scanning photoacoustic microscopy of oxygen saturation at 1-MHz A-line rate,” Photoacoustics 20, 100195 (2020). [CrossRef]
8. J. Chen, Y. Zhang, X. Li, J. Zhu, D. Li, S. Li, C.-S. Lee, and L. Wang, “Confocal visible/NIR photoacoustic microscopy of tumor with structural, functional and nanoprobe contrasts,” Photonics Res. 8(12), 1875 (2020). [CrossRef]
9. J. Ahn, J. Y. Kim, W. Choi, and C. Kim, “High-resolution functional photoacoustic monitoring of vascular dynamics in human fingers,” Photoacoustics 23, 100282 (2021). [CrossRef]
10. F. Zhong, Y. Bao, R. Chen, Q. Zhou, and S. Hu, “High-speed wide-field multi-parametric photoacoustic microscopy,” Opt. Lett. 45(10), 2756–2759 (2020). [CrossRef]
11. C. Zhang, H. Zhao, S. Xu, N. Chen, K. Li, X. Jiang, L. Liu, Z. Liu, L. Wang, K. K. Y. Wong, J. Zou, C. Liu, and L. Song, “Multiscale high-speed photoacoustic microscopy based on free-space light transmission and a MEMS scanning mirror,” Opt. Lett. 45(15), 4312 (2020). [CrossRef]
12. G. L. Semenza, “Oxygen sensing, hypoxia-inducible factors, and disease pathophysiology,” Annu. Rev. Pathol.: Mech. Dis. 9(1), 47–71 (2014). [CrossRef]
13. M. Li, Y. Tang, and J. Yao, “Photoacoustic tomography of blood oxygenation: A mini review,” Photoacoustics 10, 65–73 (2018). [CrossRef]
14. J. M. Yang, C. Favazza, R. Chen, J. Yao, X. Cai, K. Maslov, Q. Zhou, K. K. Shung, and L. V. Wang, “Simultaneous functional photoacoustic and ultrasonic endoscopy of internal organs in vivo,” Nat. Med. 18(8), 1297–1302 (2012). [CrossRef]
15. T. Wang, S. Nandy, H. S. Salehi, P. D. Kumavor, and Q. Zhu, “A low-cost photoacoustic microscopy system with a laser diode excitation,” Biomed. Opt. Express 5(9), 3053 (2014). [CrossRef]
16. M. Erfanzadeh, P. D. Kumavor, and Q. Zhu, “Laser scanning laser diode photoacoustic microscopy system,” Photoacoustics 9, 1–9 (2018). [CrossRef]
17. A. Stylogiannis, L. Prade, A. Buehler, J. Aguirre, G. Sergiadis, and V. Ntziachristos, “Continuous wave laser diodes enable fast optoacoustic imaging,” Photoacoustics 9, 31–38 (2018). [CrossRef]
18. L. Zeng, G. Liu, D. Yang, and X. Ji, “Portable optical-resolution photoacoustic microscopy with a pulsed laser diode excitation,” Appl. Phys. Lett. 102(5), 053704 (2013). [CrossRef]
19. L. Zeng, Z. Piao, S. Huang, W. Jia, and Z. Chen, “Label-free optical-resolution photoacoustic microscopy of superficial microvasculature using a compact visible laser diode excitation,” Opt. Express 23(24), 31026 (2015). [CrossRef]
20. H. Zhong, J. Zhang, T. Duan, H. Lan, M. Zhou, and F. Gao, “Enabling both time-domain and frequency-domain photoacoustic imaging by a fingertip laser diode system,” Opt. Lett. 44(8), 1988 (2019). [CrossRef]
21. X. Li, V. T. C. Tsang, L. Kang, Y. Zhang, and T. T. W. Wong, “High-speed high-resolution laser diode-based photoacoustic microscopy for in vivo microvasculature imaging,” Vis. Comput. Ind. Biomed. Art 4(1), 1 (2021). [CrossRef]
22. A. Stylogiannis, L. Riobo, L. Prade, S. Glasl, S. Klein, G. Lucidi, M. Fuchs, D. Saur, and V. Ntziachristos, “Low-cost single-point optoacoustic sensor for spectroscopic measurement of local vascular oxygenation,” Opt. Lett. 45(24), 6579 (2020). [CrossRef]
23. (ANSI) American National Standards Institute Inc, ANSI Z136.1–2007: American National Standard for Safe Use of Lasers (2007).
24. M. L. Li, J. T. Oh, X. Xie, G. Ku, W. Wang, C. Li, G. Lungu, G. Stoica, and L. V. Wang, “Simultaneous molecular and hypoxia imaging of brain tumors in vivo using spectroscopic photoacoustic tomography,” Proc. IEEE 96(3), 481–489 (2008). [CrossRef]
25. J. Laufer, C. Elwell, D. Delpy, and P. Beard, “In vitro measurements of absolute blood oxygen saturation using pulsed near-infrared photoacoustic spectroscopy: accuracy and resolution,” Phys. Med. Biol. 50(18), 4409–4428 (2005). [CrossRef]
26. K. Briely-Sabo and a. Bjornerud, “Accurate de-oxygenation of ex vivo whole blood using sodium dithionite,” Proc. Intl. Sot. Mag. Reson. Med 117(1985), 2025 (2000).
27. F. Gao, Q. Peng, X. Feng, B. Gao, and Y. Zheng, “Single-wavelength blood oxygen saturation sensing with combined optical absorption and scattering,” IEEE Sens. J. 16(7), 1943–1948 (2016). [CrossRef]
28. K. Sei, M. Fujita, T. Hirasawa, S. Okawa, T. Kushibiki, H. Sasa, K. Furuya, and M. Ishihara, “Measurement of blood-oxygen saturation using a photoacoustic technique in the rabbit hypoxemia model,” J Clin Monit Comput 33(2), 269–279 (2019). [CrossRef]
29. S. Aswini, A. Suresh, S. Priya, and B. V. Santhosh Krishna, “Retinal vessel segmentation using morphological top hat approach on diabetic retinopathy images,” Proc. 4th IEEE Int. Conf. Adv. Electr. Electron. Information, Commun. Bio-Informatics, AEEICB 2018 (2018).
30. M. Sun, C. Li, N. Chen, H. Zhao, L. Ma, C. Liu, Y. Shen, R. Lin, and X. Gong, “Full three-dimensional segmentation and quantification of tumor vessels for photoacoustic images,” Photoacoustics 20, 100212 (2020). [CrossRef]
31. E. P. Rivers, D. S. Ander, and D. Powell, “Central venous oxygen saturation monitoring in the critically ill patient,” Current Opinion in Critical Care 7(3), 204–211 (2001). [CrossRef]
32. J. Yao, L. Wang, J.-M. Yang, L. S. Gao, K. I. Maslov, L. V. Wang, C.-H. Huang, and J. Zou, “Wide-field fast-scanning photoacoustic microscopy based on a water-immersible MEMS scanning mirror,” J. Biomed. Opt. 17(8), 1 (2012). [CrossRef]
33. M. Chen, X. Duan, B. Lan, T. Vu, X. Zhu, Q. Rong, W. Yang, U. Hoffmann, J. Zou, and J. Yao, “High-speed functional photoacoustic microscopy using a water-immersible two-axis torsion-bending scanner,” Photoacoustics 24, 100309 (2021). [CrossRef]
