Open Access
Add to BibSonomy
Abstract
Multispectral optoacoustic tomography (MSOT) has become the dominant technical solution for photoacoustic imaging (PAI). However, the laser source of fiber output in the current MSOT method is typically a TEM00 Gaussian beam, which is prone to artifacts and incomplete due to the uneven distribution of the irradiated light intensity. Here, we propose a novel method to improve the quality of photoacoustic image reconstruction by modulating the wavefront shaping of the incident laser beam based on the designed scattering structure. In the experiment, we add the designed scattering structure to the current hemispherical photoacoustic transducer array device. Through experiments and simulations, we investigate and compare the effects of different scattering structures on laser intensity modulation. The results show that an ED1-C20 diffusion structure with a scattering angle of 20 degrees has the most effective modulation of the beam intensity distribution. And we choose gold nanoparticles of 50-100 nanometers (nm) diameters and index finger capillary vessels respectively as the medium of PAI. We obtain the highest ratio of PAI area increases of gold nanoparticles and index finger to devices compare without scattering structure is 29.69% and 634.94%, respectively. Experimental results demonstrate that our method is significantly higher quality than traditional methods, which has great potential for theoretical application in medical PAI.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
Retraction
This article has been retracted. Please see:Guodong Tong, Artur Luzgin, Jun Xia, Liyan Xu, Hao Zhang, Chengkun Dong, Zhihai Wu, Jun Wu, Yichen Zhang, and Peiwu Qin, "Improved photoacoustic images via wavefront shaping modulation based on the scattering structure: retraction," Opt. Express 31, 31670-31670 (2023)
https://opg.optica.org/oe/abstract.cfm?uri=oe-31-20-31670
1. Introduction
Multispectral optoacoustic tomography (MSOT) is currently a hot spot in the research of photoacoustic (PA) imaging for biological tissue [1–3]. Photoacoustic tomography (PAT) imaging is superior to traditional ultrasound imaging (UI) [4–6], Raman scattering imaging (RSI) [7,8], and photothermal imaging (PTI) [9,10]. The quality of PAI is closely related to the photoacoustic energy conversion efficiency [11,12]. The main solutions to improve the quality of PAT imaging are: contrast agent and optical field wave-front modulation. Using contrast agents to improve the absorption efficiency of light energy by the medium [13,14], such as near infrared laser imaging of biological tissue capillaries, the absorption spectra of two forms of hemoglobin (oxyhemoglobin and deoxyhemoglobin) are in the visible and near-infrared bands [15,16]. Because the absorbance of hemoglobin for the near-infrared spectrum is higher than that of other blood molecules, it has become a powerful contrast agent for vascular PAI. The other methods such as wave-front modulation of the light field to improve the photoacoustic efficiency [17,18], and selecting different laser wavelengths, different pulse repetition frequencies and light field intensity modulation, etc.
When short pulsed laser waves irradiate biological tissue, non-radiative transitions occur, which in turn generates heat and undergoes a very small expansion deformation [19–21], the pressure before thermal expansion we call the initial pressure (IP). For the PA imaging modality, the uneven distribution of the beam will cause the incompleteness of the photoacoustic signal, and the reconstruction of the PA imaging will induce the quality of image. Therefore, wave-front shaping is one way to improve the photoacoustic energy conversion efficiency and enhance the quality of photoacoustic images [22–24]. The wavelength, average power and intensity distribution of the laser source are the main targets of wave-front regulation [25–27]. The structure of the laser source illumination is also one of the methods to improve the quality of PA imaging, currently it is mainly annular illumination [28]. The illumination pattern can be classified into linear array, plane array, ring array and hemispherical array, etc [29]. In addition, the photoacoustic computer tomography (PACT) system of the reflection projection PA imaging detector also be one method to control the laser source wavefront [12,30–32]. In theory, when the detectors are distributed on the spherical surface, the viewing angle is the largest and the imaging performance is the best [33–35].
In this study, we propose a novel method to improve the PA imaging quality of MSOT by modulating the incident laser source of fiber output based on the scattering structure. First, a specific scattering structure is added to the current hemispherical transducer array structure to modulate the intensity distribution of the incident laser beam to achieve a more uniform distribution of laser intensity. A larger contact surface is used to efficiently generate photoacoustic signals, which is beneficial to enhance the quality of the reconstructed photoacoustic image. Because most of the beams emitted by pulsed lasers through optical fibers are Gaussian beams, we know that the characteristics of Gaussian beams are that the strongest part of the laser intensity is concentrated in the center, which may not only cause damage to the medium when center of the beam is strong enough, but also the laser intensity except the center of the laser beam is relatively weak, the photoacoustic signal is very weak and the PA imaging is very unclear. Our method can reduce the laser damage due to the high centroid intensity, and can achieve the PA imaging more accurately and comprehensively.
2. Method and theory
2.1 Photoacoustic images system
Here, we use localization photoacoustic tomography (LPT) with the aim of overcoming spatial resolution barriers in PA imaging and tomography in acoustic diffraction. To demonstrate true 3D localization capabilities, the PAT system was used in my current work on real-time volumes reported prior to my visit to the lab [36]. In addition, we digitized 256 time-resolved PA imaging signals detected simultaneously at 40 M samples per second using a custom data acquisition system (DAQ, Falkenstein Mikrosysteme, GmbH, Taufkirchen, Germany). The operating frequency of the laser can reach a pulse repetition frequency (PRF) of 100 Hz. The matrix array had a central opening through which we directed the beam using a custom-made fiber bundle (Ceramoptec GmbH, Bonn, Germany).
As shown in Fig. 1(a), it adopts a spherical matrix detection array, which is uniformly filled in a hemisphere by 256 independent piezoelectric composite elements with a center frequency of 4 MHz. The array was fixed upwards, used as an acoustic coupling medium, and was further used to hold a tunable (700-900 nm) optical parametric oscillator (OPO) laser. Finally, DAQ transfers the data to a PC for further analysis. Data collection is performed synchronously.
Fig. 1. (a) Spherical array of ultrasound transducers is used to acquire a three-dimensional PA imaging of flowing absorber for each laser pulse. (b)The schematic of experimental setup; (c) Pictorial view of the experimental setup inside the yellow box in (a).
Based on the heat transfer principle of photoacoustic imaging, the key of our work is to modulate the laser beam to irradiate as much as possible on the target material, improve the photoacoustic efficiency, and then provide photoacoustic imaging quality. The heat transfer is first calculated using the Gaussian formula. Given the heating function , the generation and propagation of photoacoustic wave pressure $p({\vec{r},t} )$ in an acoustically homogeneous inviscid medium is governed by
In the post-processing of photoacoustic imaging for image reconstruction, we have also optimized existing algorithms, including the commonly used interpolated matrix-model inversion (IMMI) method [37], the iterative algorithm (LSQR) [38], etc.
2.2 Signal processing and noise processing
We first bandpass filtered the raw collected photoacoustic signal to reduce noise and deconvolved it with the impulse response of the transducer element. Specifically, deconvolution was performed using a Wiener filter with a 0 signal-to-noise ratio (SNR), where the impulse response was measured by recording the resulting signal at the center of a spherical array. We applied a second-order bandpass Butterworth filter with cutoff frequencies of 0.1 and 8 MHz to the digitized signal to remove low-frequency offset and high-frequency noise. PA image reconstruction is then performed on a Graphics Processing Unit (GPU) on a Cartesian grid of 320 ×320 ×160 voxels using a three-dimension (3D) model-based algorithm. Briefly, the PA reconstruction method consists in the numerical discretization of the PA forward model, where the pressure signal at a set of points and instants can be represented in vector form as
where ${\textbf h}$ is the initial PA pressure and ${\textbf A}$ is the PA model-matrix. The reconstruction algorithm is derived from estimating the distribution of the initial PA pressure. This achieved by calculating the solution of a least square problem defined as.Equation (3) without the regularization term included. Iteratively using the LSQR method. Solve (10 iterations).
2.3 Mathematical formulation of diffusion laser intensity
We assume that incidence laser field $x({{\textbf r^{\prime}}} )$ impinge onto the diffuser, ${\textbf r^{\prime}}$ and $ {\textbf r}$ are the coordinates at input and output facets of diffusion. Following corresponding scattered field is $y({\textbf r} )$. Using a Green’s function $G({{\textbf r},{\textbf r^{\prime}}} )$ to describe the scatter diffuser related to the incident and output laser filed, the equation of the output scattered laser field is given as,
In addition, as shown in Fig. 1(a), we need mathematically formulate the intensity cross-correlation coefficient. All the scattered laser field exhibit complex Gaussian distribution. Then, according to the Reed’s moment theorem, the intensity cross-correlation coefficient can be expressed as,
In signaling, the photoacoustic signal is a nonlinear signal. Since the density and magnitude
The schematic diagram of the experiment setup is shown in Fig. 1(b). The near-infrared laser (laser model: Spitlight 1200 OPO) is a near-infrared laser with a wavelength of 800 nm. Laser pulse repetition frequency is 100 Hz, the devices also include water tank, hemispherical photoacoustic receiver, and so on. Here, photoacoustic transducer and there eight sensor wires to the photoacoustic transducer, the transducer is connected to the computer for real-time transmission photoacoustic data and recorded data. The reconstructed photoacoustic image can be visualized in real time through the computer terminal. Figure 1(c) shows the fixed position of the designed diffuser structure, the yellow boxes in Fig. 1(b) is for the experimental setup in Fig. 1(c), the hemispherical receiver device is placed in a water tank, and the water depth in the water tank submerge the hemispherical receiver structure.
The specific installation position of the scattering structure is shown in Fig. 1(b). The diameter of the upper half of the structure is 7.95 mm, which is same as the diameter of the fiber launch port. It can be insert into the fiber launch port. The designed diffuser diameter is maintained. In the experiment, diffusers with different scattering angles and specific sizes are coupled with the diffuser structure, and then the wavefront of the laser beam is modulated to achieve the scattering effect of the laser beam.
3. Experimental setup
In the experiment, considering the effect that the water in the water tank may affect the scattering structure, we first use glue to seal the diffuser structure and the optical fiber transmitter. In order to compare with a specific size and those without diffusers for comparison, the customized scattering structure includes: Beam Diffuser DG10-220, DG10-600, DG10-1500, ED1-C20 and ED1-C50 (from Thorlabs Company). Here, ED1-C20 is an engineered Diffuser with circle patter diffuser and scattering angle of 20 degrees, ED1-C50 is an engineered Diffuser with circle patter diffuser and scattering angle 50 degrees, and the Relative intensity and scatter angle (SA) of the selected scattering structures are shown in Fig. 2. We test these different models of diffuser and structures without diffuser. The wavefront shape modulation of the laser beam is performed according to the scattering properties of the diffuser in Fig. 2, and the comparison is made with the one without the scattering structure, and the structure is shown Fig. 3(a).
Fig. 2. The Relative intensity and scatter angle (SA) curve diagram of the Beam Diffuser DG10-220, DG10-600, DG10-1500, ED1-C20, ED1-C50. (Data sources from www.thorlabs.com)
Fig. 3. (a) Dimensions and three-dimensional (3D) images of the designed diffuser structures, as well as different models of the designed diffusers (Beam Diffuser DG10-220, DG10-600, DG10-1500, ED1-C20, ED1-C50). Different types of diffusers are installed on the diffuser structure, and then the diffuser structure is mounted on the semicircular optical fiber transmitter for testing. (b) Using a CCD camera (Balser acA1440-220uc, made in Germany) to capture the light intensity distribution after wavefront shape modulation of the incident laser source by different diffusers. (c) Distance comparison of laser beam intensity full width at half maximum (FWHM) between different types of the designed diffusers and without diffuser modulation.
Before performing the photoacoustic experiments, we first use a CCD camera (Balser acA1440-220uc, made in Germany) to photograph the intensity of the laser beam, and use the Pylon Viewer (version 6.2.0.21487) software to capture the light intensity image in real time. Then we compare the degree of dispersion of the modulation of the laser beam intensity by different designed diffusers. As shown in Fig. 3(b), the laser beam intensity distribution varies from no scattering structure to designed diffuser DG10-220, diffuser DG10-600, DG10-1500, ED1-C20 and ED1-C50. According to the corresponding attributes of different diffusers in Fig. 2, the laser beam intensity distribution ranges are different. In Fig. 3(b) and Fig. 4, by comparing the experimental results, we found that the diffuser model ED1-C20 has the best dispersion effect on the wavefront shape modulation of the laser beam. The laser beam intensity structure diagram obtained by MATLAB software analysis is shown in Fig. 4, and calculate their corresponding FWHM size and laser beam intensity. The FWHM of various scattering structures are shown in Fig. 4.
Fig. 4. (a) The normalized image of the intensity distribution after different types of diffusers modulating the incident laser source; (b) Normalized laser intensity distribution curves after modulating the incident laser source with different scattering structures.
By comparing the normalized laser beam intensity distribution and curve distribution in Fig. 4, it is obvious that ED1-C20 has the best modulation effect on the laser beam field. In Fig. 4(b), the maximum laser field intensity of ED1-C20 (Max = 0.8) is the largest. Although the distribution of ED1-C50 is more uniform, the overall laser field intensity is weaker than the ED1-C20. Therefore, we select the best ED1-C20 for the experiment, and compare the photoacoustic image with the traditional method (No diffuser).
4. Results and discussion
Through the above experiments and simulation analysis, in contrast to other types of scattering structures, the diffuser of the ED1-C20 type has an ideal wavefront shape modulation effect on the laser beam, and the laser beam intensity distribution is uniform and intensity is the strongest. Normalized intensity value (Max = 0.8) of the ED1-C20 diffuser is greater than the laser beam intensity of other types. Therefore, for the verification part of the photoacoustic experiment, we use the ED1-C20 diffuser and the scattering structure without diffuser for experiment and comparison.
In the experiment, we put the gold-nanoparticles into agar, the diameter of the gold nanoparticles is about 50-100 nanometers (nm), and then smear the agar containing gold nanoparticles on the glass slide. The photoacoustic images obtained after the experiment are shown in Fig. 5.
Fig. 5. Gold-nanoparticles photoacoustic imaging with or without Diffusion modulation. (a) with diffusion modulation device; (b) Without diffusion modulation device. (c) Comparison of the SNR of photoacoustic signals generated by gold nanoparticles with scattering structure at different wavelengths and without scattering structure (d) Comparison of binary grayscale images of photoacoustic images; (e) Binary grayscale PA imaging area difference between the scattering modulation device and without scattering modulation device.
Here, we use the View4D toolbox of MATLAB software to process the received photoacoustic signals and analyze the generated photoacoustic images, the pixel size of two photoacoustic images is $584 \times 584$. And we use the maximum intensity projection (MIP) technology to process the photoacoustic images and select the transversal images as the key research. Then we lock the photoacoustic images at the same position for comparison. By testing the ratio of the light beams generated by gold nanoparticles at different wavelengths, we compared the results with and without the scattering structure. From Fig. 5(c), it can be concluded that the intensity of the photoacoustic signal we generate is much greater than the signal generated by noise. Therefore, the reconstructed images in Fig. (a) and (b) can prove to be images of gold nanoparticles. The white circle marked in Fig. 5 are the areas with obvious differences. Hereby, we calculated that area of pixel values of the photoacoustic signal images by using the grayscale image binarization algorithm. The area pixel value of the photoacoustic signal image in Fig. 5(a) is 3840, and the area of pixel value of the photoacoustic signal image in Fig. 5(b) is 2981. The difference between the two is 193367. Therefore, the ratio of the increase value of the PA imaging area to the value of Fig. 5(b) is 29.69%, which clearly shows that images without the scattering structure is not clear enough, not complete enough. Photoacoustic images with scattering structures show more comprehensive imaging of gold nanoparticles. In addition, we use the readBinary and sum_White functions of the MATLAB software to capture and compare the bright spot areas of gold nanoparticles of photoacoustic images, and subtract the bright spot area curves of the two images, as shown Fig. 5(d). They are proved that the scattering structure is beneficial to enhance the clarity and accuracy of the photoacoustic image.
According to the experimental setup shown in Fig. 1, we put index finger in the photoacoustic receiver in the water tank, and by moving the index finger slowly, the laser beam irradiates different positions of the index finger, and the light at different positions of the index finger is acquired and recorded through the photoacoustic signal receiver. The received photoacoustic signals are processed by the photoacoustic transducer and MATLAB software. Here, we also use the View4D toolbox of MATLAB software to process the received photoacoustic signals and analyze the generated photoacoustic images, the pixel size of two photoacoustic images is $584 \times 584$. And we use MIP technology to process the photoacoustic images and select the transversal images for comparison. In this study, we lock the three same positions of the index finger for analysis. As shown Fig. 6, three transversal images of different index finger positions are selected with the single Frame = 4 in View4D toolbox, and the position of Time-Activity curve (value = 127.120.29). Figure 6 shows the big difference between the part indicated by the white circles (no diffuser) and black circles (withdiffuser). Hereby, Figs. 6(h) and (i) are chosen as examples for study comparison. First, we calculated the pixel value of the PA imaging area by using the grayscale image binarization algorithm, the pixel value of the PA imaging area in Fig. 6(h) is 339855, and pixel value of the PA imaging area in Fig. 6(k) is 146488, and the difference is 193367, therefore, the ratio of the difference to Fig. 6(k) is 132%; The pixel value of the PA imaging area in Fig. 6(i) is 302294, and the pixel value of the PA imaging area in Fig. 6(m) is 41132, and the difference is 261162, the result of the ratio of the difference to Fig. 6(m) is 634.94%. In short, the results of the above differences obviously present that the photoacoustic images (Fig. 6(h) and (i) with the scattering structure show a more comprehensive and accurate display effect than the photoacoustic images (Fig. 6(k) and (m)) without the scattering structure. In addition, Fig. 6 (n) and (o) show the subtraction difference values images of the brightness area curves of Fig. 6(h) minus (k), (i) minus (m), respectively. In the Fig. 6(p) and (q), the red part shows the more photoacoustic image display than the non-scattering structure, and the blue and hades represent the overlap. The parts marked by the white arrows demonstrate that the image after scattering shape modulation by this method can perform photoacoustic imaging with higher quality than the image non-scattering structure. According to the above experimental results and numerical analysis, the photoacoustic images display generated by our proposed method are more complete and clearer. This greatly improves the accuracy of photoacoustic images and its application value.
Fig. 6. Capillary PA imaging of index finger compared with and without diffusion device. (a) - (c) and (g) - (i) with diffusion modulation device; (d) - (f) and (i) - (m) without diffusion device. (n) Binary grayscale PA imaging area value subtraction image of the diffusion modulation device (h) minus the without diffusion device (k). (o) Binary grayscale PA imaging area value subtraction image of the diffusion modulation device (i) minus without the diffusion device (m). (p) and (q) The difference image of the non-scattering structure is subtracted after modulating the scattering corresponding to (n) and (0), respectively. Red images represent the more photoacoustic image display than the non-scattering structure, and the blue and hades represent the overlap.
5. Conclusion
In this study, we propose an advanced method to enhance the quality of PA imaging. Based on the traditional PA imaging method, we designed a specific ED1-C20 diffuser scattering structure with a scattering angle of 20 degrees and diameter approximately equal to the diameter of the optical fiber transmitter, which modulates the wavefront of the incident initial Gaussian beam to increase the irradiation area of the beam. In the experiment, we used 50-100 nm diameters gold nanoparticles and the index finger as object medium, respectively. The SNR we measured with the scattering structure is also about 10 dB larger than the SNR without the scattering structure. The ratio of the increased pixel value of the PA imaging area of gold nanoparticles and index finger to devices with scattering structures is the highest at 29.69% and 634.94%, respectively. Comparison with photoacoustic images without scattering structures, the PA imaging of the whole area is more obvious, and the display of the whole area is still clear enough. In addition, we use the View4D toolbox capture the transverse images of photoacoustic images, and use the readBinary and sum_White functions of the MATLAB software to compare the brightness areas curves difference of photoacoustic images. The experimental and numerical results demonstrate that our method not only improves the overall image integrity, but also displays the photoacoustic image more accurately. It will have more research value for medical imaging and PA imaging.
Funding
National Key Research and Development Program of China (2021YFF0701100, 2021YFB3600502); National Natural Science Foundation of China (62075040).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. L. R. McNally, M. Mezera, D. E. Morgan, P. J. Frederick, E. S. Yang, I.-E. Eltoum, and W. E. Grizzle, “Current and emerging clinical applications of multispectral optoacoustic tomography (MSOT) in oncology,” Clin. cancer Res. 22(14), 3432–3439 (2016). [CrossRef]
2. D. Razansky, A. Buehler, and V. Ntziachristos, “Volumetric real-time multispectral optoacoustic tomography of biomarkers,” Nat. Protoc. 6(8), 1121–1129 (2011). [CrossRef]
3. A. Taruttis and V. Ntziachristos, “Advances in real-time multispectral optoacoustic imaging and its applications,” Nat. Photonics 9(4), 219–227 (2015). [CrossRef]
4. R. Manwar, M. Zafar, and Q. Xu, “Signal and image processing in biomedical photoacoustic imaging: a review,” Optics 2(1), 1–24 (2020). [CrossRef]
5. J. Jose, R. G. H. Willemink, W. Steenbergen, C. H. Slump, T. G. van Leeuwen, and S. Manohar, “Speed-of-sound compensated photoacoustic tomography for accurate imaging,” Med. Phys. 39(12), 7262–7271 (2012). [CrossRef]
6. J. K. Gamelin, A. Aquirre, A. Maurudis, F. Huang, D. Castillo, L. V. Wang, and Q. Zhu, “Curved array photoacoustic tomographic system for small animal imaging,” J. Biomed. Opt. 13(2), 024007 (2008). [CrossRef]
7. H. Feng, Q. Fu, W. Du, R. Zhu, X. Ge, C. Wang, Q. Li, L. Su, H. Yang, and J. Song, “Quantitative assessment of copper (II) in Wilson’s disease based on photoacoustic imaging and ratiometric surface-enhanced Raman scattering,” ACS Nano 15(2), 3402–3414 (2021). [CrossRef]
8. V. Neuschmelting, S. Harmsen, N. Beziere, H. Lockau, H.-T. Hsu, R. Huang, D. Razansky, V. Ntziachristos, and M. F. Kircher, “Dual-modality surface-enhanced resonance Raman scattering and multispectral optoacoustic tomography nanoparticle approach for brain tumor delineation,” Small 14(23), 1800740 (2018). [CrossRef]
9. X. Zhen, K. Pu, and X. Jiang, “Photoacoustic Imaging and Photothermal Therapy of Semiconducting Polymer Nanoparticles: Signal Amplification and Second Near-Infrared Construction,” Small 17(6), 2004723 (2021). [CrossRef]
10. Y. Lyu, J. Zeng, Y. Jiang, X. Zhen, T. Wang, S. Qiu, X. Lou, M. Gao, and K. Pu, “Enhancing both biodegradability and efficacy of semiconducting polymer nanoparticles for photoacoustic imaging and photothermal therapy,” ACS Nano 12(2), 1801–1810 (2018). [CrossRef]
11. Y. Mantri and J. V. Jokerst, “Engineering plasmonic nanoparticles for enhanced photoacoustic imaging,” ACS Nano 14(8), 9408–9422 (2020). [CrossRef]
12. A. Fatima, K. Kratkiewicz, R. Manwar, M. Zafar, R. Zhang, B. Huang, N. Dadashzadeh, J. Xia, and K. M. Avanaki, “Review of cost reduction methods in photoacoustic computed tomography,” Photoacoustics 15, 100137 (2019). [CrossRef]
13. E. V. Shashkov, M. Everts, E. I. Galanzha, and V. P. Zharov, “Quantum dots as multimodal photoacoustic and photothermal contrast agents,” Nano Lett. 8(11), 3953–3958 (2008). [CrossRef]
14. J. F. Lovell, C. S. Jin, E. Huynh, H. Jin, C. Kim, J. L. Rubinstein, W. C. W. Chan, W. Cao, L. V. Wang, and G. Zheng, “Porphysome nanovesicles generated by porphyrin bilayers for use as multimodal biophotonic contrast agents,” Nat. Mater. 10(4), 324–332 (2011). [CrossRef]
15. H. F. Zhang, K. Maslov, M. Sivaramakrishnan, G. Stoica, and L. V. Wang, “Imaging of hemoglobin oxygen saturation variations in single vessels in vivo using photoacoustic microscopy,” Appl. Phys. Lett. 90(5), 053901 (2007). [CrossRef]
16. J. A. Guggenheim, T. J. Allen, A. Plumb, E. Z. Zhang, M. Rodriguez-Justo, S. Punwani, and P. C. Beard, “Photoacoustic imaging of human lymph nodes with endogenous lipid and hemoglobin contrast,” J. Biomed. Opt. 20(05), 1 (2015). [CrossRef]
17. R. Horstmeyer, H. Ruan, and C. Yang, “Guidestar-assisted wavefront-shaping methods for focusing light into biological tissue,” Nat. Photonics 9(9), 563–571 (2015). [CrossRef]
18. R. Gao, Z. Xu, L. Song, and C. Liu, “Breaking Acoustic Limit of Optical Focusing Using Photoacoustic-Guided Wavefront Shaping,” Laser & Photonics Reviews 15(8), 2000594 (2021). [CrossRef]
19. F. A. McDonald, “Practical quantitative theory of photoacoustic pulse generation,” Appl. Phys. Lett. 54(16), 1504–1506 (1989). [CrossRef]
20. R. S. Quimby and W. M. Yen, “Photoacoustic theory including energy migration,” J. Appl. Phys. 51(9), 4985–4989 (1980). [CrossRef]
21. F. A. McDonald and G. C. Wetsel Jr, “Generalized theory of the photoacoustic effect,” J. Appl. Phys. 49(4), 2313–2322 (1978). [CrossRef]
22. O. Tzang and R. Piestun, “Lock-in detection of photoacoustic feedback signal for focusing through scattering media using wave-front shaping,” Opt. Express 24(24), 28122–28130 (2016). [CrossRef]
23. Y. Guo, B. Li, and X. Yin, “Single-shot compressed photoacoustic tomographic imaging with a single detector in a scattering medium,” Phys. Rev. Appl. 13(4), 044009 (2020). [CrossRef]
24. T. W. Murray, M. Haltmeier, T. Berer, E. Leiss-Holzinger, and P. Burgholzer, “Super-resolution photoacoustic microscopy using blind structured illumination,” Optica 4(1), 17–22 (2017). [CrossRef]
25. S. Tuohy and A. G. Podoleanu, “Depth-resolved wavefront aberrations using a coherence-gated Shack-Hartmann wavefront sensor,” Opt. Express 18(4), 3458–3476 (2010). [CrossRef]
26. R. Aviles-Espinosa, J. Andilla, R. Porcar-Guezenec, O. E. Olarte, M. Nieto, X. Levecq, D. Artigas, and P. Loza-Alvarez, “Measurement and correction of in vivo sample aberrations employing a nonlinear guide-star in two-photon excited fluorescence microscopy,” Biomed. Opt. Express 2(11), 3135–3149 (2011). [CrossRef]
27. X. Tao, B. Fernandez, O. Azucena, M. Fu, D. Garcia, Y. Zuo, D. C. Chen, and J. Kubby, “Adaptive optics confocal microscopy using direct wavefront sensing,” Opt. Lett. 36(7), 1062–1064 (2011). [CrossRef]
28. C. Cai, X. Wang, K. Si, J. Qian, J. Luo, and C. Ma, “Feature coupling photoacoustic computed tomography for joint reconstruction of initial pressure and sound speed in vivo,” Biomed. Opt. Express 10(7), 3447–3462 (2019). [CrossRef]
29. Y. Zhou, J. Yao, and L. V. Wang, “Tutorial on photoacoustic tomography,” J. Biomed. Opt. 21(6), 061007 (2016). [CrossRef]
30. Y. Xu, L. V. Wang, G. Ambartsoumian, and P. Kuchment, “Reconstructions in limited-view thermoacoustic tomography,” Med. Phys. 31(4), 724–733 (2004). [CrossRef]
31. J. Xia, J. Yao, and L. V Wang, “Photoacoustic tomography: principles and advances,” Electromagn. waves (Cambridge, Mass.) 147, 1 (2014).
32. N. Nyayapathi and J. Xia, “Photoacoustic imaging of breast cancer: a mini review of system design and image features,” J. Biomed. Opt. 24(12), 1 (2019). [CrossRef]
33. S. Na, X. Yuan, L. Lin, J. Isla, D. Garrett, and L. V. Wang, “Transcranial photoacoustic computed tomography based on a layered back-projection method,” Photoacoustics 20, 100213 (2020). [CrossRef]
34. C. Tian, C. Zhang, H. Zhang, D. Xie, and Y. Jin, “Spatial resolution in photoacoustic computed tomography,” Rep. Prog. Phys. 84(3), 036701 (2021). [CrossRef]
35. L. V. Wang and L. Gao, “Photoacoustic microscopy and computed tomography: from bench to bedside,” Annu. Rev. Biomed. Eng. 16(1), 155–185 (2014). [CrossRef]
36. T. Jetzfellner, D. Razansky, A. Rosenthal, R. Schulz, K.-H. Englmeier, and V. Ntziachristos, “Performance of iterative optoacoustic tomography with experimental data,” Appl. Phys. Lett. 95(1), 013703 (2009). [CrossRef]
37. C. C. Paige and M. A. Saunders, “LSQR: An algorithm for sparse linear equations and sparse least squares,” ACM Trans. Math. Softw. 8(1), 43–71 (1982). [CrossRef]
