This PDF file contains the front matter associated with SPIE Proceedings Volume 12631, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

Photoacoustic remote sensing has been recently developed as an all-optical, non-contact, and label-free imaging modality capable of imaging a variety of endogenous contrast agents through the detection of reflectivity modulations. Initial work described an elasto-optic refractive index perturbation model to explain reflectivity modulations observed, however, in practice the magnitude of these reflectivity modulations has been found to be orders of magnitude smaller than those typically observed experimentally. In this report we utilize a ten million frames-per-second camera to further investigate these reflectivity modulations, while also exploring other potential mechanisms of laser pulse-induced reflectivity modulations. Laser-induced motion is demonstrated both laterally for gold wires suspended and submerged in air and water, respectively, and carbon fibers submerged in water, and axial motion is observed in gold wires submerged in a depth gradient of intralipid solution. This laser-induced sample motion is anticipated to cause reflectivity modulations local to the interrogation beam profile in microscopy set-ups. Non-motion-based maximum intensity modulations of 3% are also observed in gold wires submerged in water, indicating the presence of the originally predicted reflectivity modulations. Overall, these observations are important as they provide a widefield view of laser-pulse interactions unavailable in previous point scanning-based photoacoustic remote sensing microscopy configurations, where observed mechanisms occur on time-scales orders of magnitude faster than equivalent field of view point scanning capabilities.

In conventional optoacoustic microscopy, nanosecond pulse duration lasers are employed. When a laser delivering shorter pulse durations is used, it is expected that, from a theoretical point of view, broader, higher-frequency acoustic waves to be generated, therefore a better axial resolution of the instrument. In the present report, this advantage, offered by a picosecond duration pulse laser, to experimentally demonstrate that the axial resolution of an optoacoustic microscopy instrument can be enhanced was exploited. In comparison to a 2 ns pulse duration, an improvement in the axial resolution of ~50% is demonstrated by using excitations with pulses of duration ⪅100 ps. Details of an optoacoustic microscopy instrument, operating at 532 nm, capable to provide high-resolution axial and lateral optoacoustic images, are also presented. The capabilities of the instrument are demonstrated by in-vivo images of Xenopus laevis brain with a similar ~ 3.8 μm lateral resolution throughout the whole axial imaging range.

In Photoacoustic Tomography (PAT), the aim is to estimate the initial pressure distribution based on measured ultrasound data. While several approaches utilizing deep learning for PAT have been proposed, many of these do not provide estimates on the reliability of the reconstruction. In this work, we propose a deep learning approach for the Bayesian inverse problem for PAT based on the uncertainty quantification variational autoencoder. The approach enables simultaneous image reconstruction and reliability estimation.

The frequency domain Optoacoustic (OA) wave equation is inherently inhomogeneous. The first inhomogeneous term arises because of the OA effect (i.e., conversion of optical energy into acoustical energy). The second term appears due to sound-speed mismatch between the source and the ambient medium. The conventional Green’s function method works well in absence of the second term (i.e., acoustically homogeneous source). Recently, it has been shown that a Modified Green’s Function (MGF) approach provides faithful solution to the OA wave equation for an acoustically inhomogeneous source. Herein, we employ the MGF technique for accurate estimation of the OA spectra for normal and pathological red blood cells (RBCs). The shapes in 2D mimicking normal RBC, stomatocyte and echinocyte (with six equidistant identical spicules) were simulated (with constant area ≈ 16.5 μm² ) and subsequently, the OA spectra were computed over 10- 1000 MHz by evaluating the integral equation employing the Monte Carlo integration method. The OA spectrum for an equivalent disc was also calculated for comparison. The sound-speed within the source region was taken as 1639 m/s and that of the surrounding medium was chosen as 1500 m/s. The first minimum of the OA spectra for disc and echinocyte appeared almost at the same location (440 MHz) when probed from an angle, θ=π/4 with respect to the axis of symmetry. The locations for the first minimum became 280 and 390 MHz, respectively for normal RBC and stomatocyte (for θ=π/4). These 0A spectral features may be useful for morphological characterization of normal and deformed RBCs.

Photoacoustic Tomography (PAT) is a useful tool for fast 3D imaging that provides structural, molecular, and functional in vivo information. It is capable of producing 3D images using a multi-element hemispherical array transducer. PAT images can be enhanced a great number of ultrasonic transducer components with multiplexers, but this can result in high costs and slow temporal resolution because of using multiplexers. In this research, we present a deep learning solution to improve both the spatial and temporal resolution in PAT. We demonstrated that the trained neural network enhanced the image quality of a quarter-cluster-sampled data of static whole-body imaging. Our approach increased limited-view aperture and the spatial resolution by around three and two times, respectively. Additionally, it allowed to improve temporal resolution by four times without multiplexing. Our method also demonstrated excellent performance in contrast-enhanced PA imaging, enabling molecular imaging. Our strategy has the potential to enable high spatial and temporal resolution observation of biodynamics in 3D PAT without being limited by hardware constraints.

Photoacoustic imaging based on projection of Hadamard patterns is known to produce optical-resolution images with high contrast at low levels of radiant exposure. In this work, the requirements regarding noise level, a two-step reconstruction that reduces artifacts created by a focused detector and approaches to reduce the amount of acquired data without loss of image quality are discussed.

Ultraviolet photoacoustic remote sensing microscopy provides label-free optical absorption contrast comparable to hematoxylin staining. This has been combined with 266 nm optical scattering microscopy offering eosin-like contrast. Here, we use unsupervised deep learning-based style transfer using the CycleGAN approach to render these pseudo-colored virtual histological images in a realistic stain style comparable to the H&E gold standard in unstained human and murine tissue specimens. A multi-pathologist diagnostic concordance study found a sensitivity of 89%, specificity of 91%, and accuracy of 90%. A blinded subjective stain quality survey suggested virtual histology was preferred over frozen sections at the 95% confidence level.

Quantitative photoacoustic tomography aims at estimating optical parameters of tissue utilizing the photoacoustic effect. In this work, spectral optical parameters are estimated in one-step from photoacoustic data utilizing a Bayesian approach to inverse problems. Forward model is constructed by presenting optical absorption and scattering by their spectral representations, and by combining the models for light and ultrasound propagation. The proposed methodology is evaluated with numerical simulations in full view and limited view geometries.

A common approach in Photoacoustic Imaging (PAI) is to use a linear or curved piezoelectric transducer array, which provides flexibility and versatility during image acquisition. However, these PAI systems often have limited Field-of-View (FOV), resolution, and contrast, resulting in low quality images. In this study, a multi-transducer approach is proposed to improve FOV, resolution, and contrast, with the goal of facilitating human carotid plaque imaging. A prototype consisting of multiple Capacitive Micromachined Ultrasonic Transducers (CMUTs) on a flexible array with shared channels was developed and evaluated using simulated and ex-vivo human carotid plaque samples. In numerical simulations, the results are evaluated based on input ground truth parameters. For ex-vivo plaque samples, results for multi-transducer are evaluated and compared to the images acquired with single transducer. All the results demonstrate that the proposed approach improves contrast, FOV, and most notably, it allows resolving the structural information in the medium where more than 25% improvement in gCNR values is achieved in both simulations and experiments compared to the PA images obtained with single transducer.

Photoacoustic (PA) imaging using pump-probe excitation has been shown to detect fluorophores with high specificity. By varying the delay between the two excitation pulses and their wavelengths, the PA signal amplitude can be modulated, yielding three different and fluorophore-specific contrast mechanisms. In this study, PA images were obtained in tissue phantoms containing purified protein solutions using a ring array tomograph and a dual-OPO laser system. The results show that pump-probe excitation can detect multiple fluorescent reporters simultaneously. The method has the potential to recover biophysical parameters, such as pH and ion concentrations.

A new ultrasound-detection technology is developed for ultrahigh-resolution optoacoustic tomography and is experimentally demonstrated with bandwidths exceeding 200 MHz and lateral resolutions beyond 20 μm. Our technology is based on an optical resonator fabricated in a silicon-photonics platform, which is coated by a sensitivity-enhancing polymer, which also eliminates the parasitic effect of surface acoustic waves. Further improvement in sensitivity is achieved by a low-noise interferometric setup, which eliminates the effect of laser frequency noise on the measurement. In vivo optoacoustic tomography is performed on a mouse ear, revealing its vasculature at detail that has been previously reserved to optoacoustic microscopy.

One of the most widespread all-optical photoacoustic imaging techniques is based on Fabry-Pérot cavities with a thin polymer spacer. However, the deposition process can lead to inhomogeneities in the thickness of this sensing layer. They can be compensated by precisely controlling the interrogation beam, to provide optimal sensitivity. This is commonly achieved with slowly tunable narrowband lasers, eventually limiting the acquisition speed. We propose instead to use a broadband amplified spontaneous emission source and a fast tunable acousto-optic filter to adjust the interrogation wavelength at each pixel within a few microseconds. This enables us to maximize the sensitivity of the optical interferometric ultrasound detection at each point of the Fabry-Perot cavity. We experimentally show that this greatly enhances the detection bandwidth of the ultrasound sensors. We demonstrate the validity of this approach by performing photoacoustic imaging with a highly inhomogeneous sensor.

An all-optical, forward-viewing, optical-resolution photoacoustic endomicroscopy probe was developed for guiding minimally invasive procedures. The probe comprises a multimode fibre for the delivery of excitation laser via wavefront shaping, and a fibre-optic ultrasound sensor based on a plane-concave microresonator at the tip of a single-mode fibre. High-resolution photoacoustic microscopy images of mouse red blood cells and mouse ear vasculature were acquired, and the high scalability of the probe in terms of field-of-view and spatial resolution was demonstrated. The ultrathin photoacoustic endomicroscopy probe promises to guide minimally invasive surgery by providing both molecular and microstructural information.

Fabry-Perot (FP) sensors are typically read out using a raster scan to acquire tomographic Photoacoustic (PA) images. To speed up the recording time, wide-field illumination of the sensor in combination with a camera as detector can be used. In this study, an sCMOS camera and wavelengths around 517 nm are used to interrogate a FP sensor with a homogeneous optical thickness over a 4 cm² aperture. The recorded time series show PA signals are acquired over the entire area of the interrogation beam. The performance of the system, such as the noise equivalent pressure, is evaluated.

Biomedical photoacoustics is usually used to image absorption-based contrast in soft tissues up to depths of several centimeters and with sub-millimeter resolution. By contrast, measuring Photoacoustic (PA) signals through hard bone tissue shows severe signal degradation due to aberration and high attenuation of high frequency acoustic signal components. This is particularly noticeable when measuring through thicker, human, skull bone. Which is the main reason why transcranial PA imaging in humans has so far proved challenging to implement. To tackle this challenge, we developed an optical resonator sensor based on a previous planar-concave design. This sensor was found to be highly suitable for measuring the low-pressure amplitude and low acoustic frequency signals that are transmitted through human cranial bone. A plano-concave optical resonator sensor was fabricated to provide high sensitivity in the acoustic frequency range of DC to around 2 MHz, a low noise equivalent pressure and a small active element size enabling it to significantly outperform conventional piezoelectric transducers when measuring PA waves transmitted through ex vivo human cranial bones.

Miniaturising ultrasonic field mapping systems could lead to novel endoscopes capable of photoacoustic tomography and other techniques. However, developing high-resolution arrays of sensitive, sub-millimetre scale ultrasound sensors presents a challenge for traditional piezoelectric transducers. To address this challenge, we conceived an ultrasonic detection concept in which an optical ultrasonic sensor array is read out using a laser beam scanned through a 0.24 mm diameter multimode optical fibre using optical wavefront shaping. We demonstrate this system enables ultrasonic field mapping with ⪆2500 measurement points, paving the way to developing miniaturized photoacoustic endoscopes and other ultrasonic systems based on the presented concept.

Silicon-photonics is a new ultrasound-detection technology, based on optical resonators, with unparalleled miniaturization levels, sensitivities, bandwidths, and capable of producing dense resonator arrays. Conventional techniques, based on tuning a continuous-wave laser to the resonator wavelength, are not scalable due to the wavelength disparity between the resonators, requiring a separate laser for each resonator. In this work, we show that also the Q-factor and transmission peak of silicon-based resonators can be pressure sensitive, develop a readout scheme based on monitoring the amplitude transmission, and demonstrate its compatibility with optoacoustic tomography.

Photoacoustic Tomography (PAT) systems based on Fabry-Perot (FP) sensors provide high-resolution images limited by the system’s sensitivity. The sensitivity is limited by the optical Q-factor of the FP cavity (i.e., the optical confinement of the interrogation laser beam in the FP cavity). In existing systems, a focused Gaussian beam is used to interrogate the sensor. While providing a small acoustic element required for high-resolution imaging, this interrogation beam naturally diverges inside the FP cavity, leading to the current sensitivity limit. To break this limit, a new approach of interrogating the FP sensor using a Bessel beam is investigated. The Noise Equivalent Pressure (NEP) and both axial and lateral PAT resolutions using Bessel beam interrogation were quantified. Bessel beam interrogation provided lower NEP, similar axial resolution, but lower lateral resolution. Thus, Bessel beam might be an alternative interrogation scheme for deep PAT imaging as high sensitivity is needed and the lateral resolution is limited by the aperture of the PAT system.

Publisher's Note: This paper, originally published on 11 August 2023, was withdrawn on 21 August 2023 per author request.

The Born series methods, namely, Traditional Born Series (TBS) and Convergent Born Series (CBS), have been recently implemented to numerically solve the time-independent Photoacoustic (PA) wave equation for an acoustically inhomogeneous source. The TBS algorithm diverges when the sound-speed mismatch is ⪆20%, however, the CBS technique provides faithful result even beyond this limit. These protocols are iterative in nature and computationally expensive. Currently, MATLAB-based implementations are available to study PA emission from a single source mimicking a cell. However, efficient numerical implementation strategy is further needed particularly to calculate PA field from a tissue. Therefore, to develop insights, uniprocessor-based C codes were realized for these schemes. The PA field (in 2D) was computed at a distance 35 μm from a source (a light absorbing disc of radius 7.5 μm) over a frequency range from 7.32 to 512 MHz with a resolution of 7.32 MHz. The sound-speed within the source region was varied from vs = 1200, 1500 and 1800 m/s, but the same quantity for the ambient medium was fixed to v_f = 1500 m/s. The C program was found to be at least ten times slower than the corresponding MATLAB program. It may be because MATLAB inherently implements parallel computing while evaluating the forward and backward Fast Fourier Transforms (FFTs) associated with the Born series approaches. Multiprocessor-based FFTs and parallel nested loops are being incorporated into the C program for enhancement of its execution speed.

The rapid localization of blood vessels in patients is important in various clinical applications, such as catheterization procedures. While optical techniques, including visual inspection, are limited in their effectiveness at depths below 1 mm, ultrasound and optoacoustic tomography can be used at deeper depths but require a spacer between the tissue and transducer to visualize superficial structures. In this work, we introduce a portable hand-held optoacoustic system that is capable of localizing blood vessels from the point of contact to a depth of 1 cm without the need for a spacer. The probe features a flat, lens-free ultrasound array which enables a largely depth independent response, though at the cost of reduced elevational resolution. In contrast to lens-based probes, where acoustic signals from outside the focal region are distorted, the amplitude of the signal from our probe only varies with depth, resulting in an imaging quality that is largely depth-independent within the imaged region. Additionally, to facilitate miniaturization, dark-field illumination is used, whereby light scattering from the tissue is exploited to homogenize the sensitivity field.

Fabry-Pèrot (FP) interferometer sensors enable highly sensitive backward mode acoustic detection in Photoacoustic (PA) imaging. They are transparent to the excitation wavelength, can be placed directly next to the PA source, and offer a broadband frequency response and high acoustic sensitivity. PA tomography using parallelized detection requires high spatial uniformity of the optical and acoustic properties, which can be hampered by contaminations during fabrication that lead to the formation of inhomogeneities and artefacts. The quality and homogeneity of the dielectric and polymer layers have a direct effect on the maximum optical phase sensitivity, and hence acoustic sensitivity. In this study, cross-sectional images of FP sensors were obtained using focused ion beam milling and ultramicrotomy followed by Scanning Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM) to evaluate different fabrication methods.

All-Optical Ultrasound (OpUS) has emerged as an imaging paradigm well-suited for minimally invasive procedures. In particular, OpUS has demonstrated potential in endovascular imaging due to its high degree of miniaturization and mechanical flexibility, high imaging resolution and immunity to electromagnetic interference. Here, we present the first human thrombus imaging using an OpUS device, which was performed on an extracted clot. The results demonstrate the feasibility of using OpUS for thrombus imaging, with the ultimate goal of guiding minimally invasive endovascular clot retrieval procedures.

In this work we give an overview and discussion of pros and cons of currently used reconstruction algorithms suited for the reconstruction of the projections of the initial pressure distribution from spatially recorded wave pattern projection images. Especially, the methods are compared in terms of reconstruction speed, the applicability in case of limited data and inhomogeneous wave propagation properties.