KEYWORDS: Education and training, Data modeling, Gallium nitride, Image restoration, Nd:YAG lasers, Deep learning, Data acquisition, Prostate, Photoacoustic spectroscopy, Laser systems engineering
Photoacoustic imaging (PAI) has emerged as a promising technique for various image guidance procedures. While convolutional neural networks (CNNs) trained on simulated radiofrequency (RF) data have been employed for point source reconstruction, their performance on real data remains a challenge. This paper addresses this limitation by introducing a novel deep learning-based method that utilizes a limited amount of experimental laser-diode-based data for the reconstruction of multiple point sources. The proposed approach employs a dual generative adversarial network (Dual-GAN) trained on experimental RF data from a combination of point source images. The Dual-GAN exhibits superior performance compared to the conventional delay-and-sum (DAS) method, demonstrating enhanced image contrast and a reduced full width at half maximum (FWHM). Notably, the axial and lateral localization errors of the Dual-GAN predictions surpass previous studies, measuring 0.028±0.018mm and 0.087±0.096mm, respectively. Additionally, the model demonstrates generalization capability by successfully reconstructing multiple point sources imaged using a different Nd:YAG laser system. This innovative method marks a significant advancement, offering improved accuracy and versatility in PAI applications involving multiple point sources.
Real-time 3-D photoacoustic (PA) imaging plays a significant role in volumetric imaging applications, such as breast imaging where PA has demonstrated significant potential. Challenges in 3-D PA imaging include long data acquisition time and limited compatibility with commonly used data acquisition systems. This paper introduces a new real-time 3-D PA data acquisition system using a matrix array transducer. Furthermore, we present a 3-D Delay and Glow (DAG) method for source localization that extends upon recently developed 2-D DAG. The experimental results show the functionality of the 3-D PA system. The DAG outperformed the conventional delay and sum (DAS) where axial, lateral, and elevational resolutions, respectively, are 0.06±0.00, 0.25±0.15, and 0.24±0.18mm for DAG and 0.14±0.06, 3.87±0.30, and 2.81±0.62mm for DAS.
Nerve imaging in radical prostatectomy would facilitate nerve-sparing surgery. Nerve imaging can be performed with photoacoustic (PA) imaging and voltage-sensitive dyes (VSD), as demonstrated in [Kang, 2019] for brain neural activity. Continuous-wave (CW) PA was used to image dynamic targets, e.g., flow [Zhao, 2021], by firing multiple modulated exciters at the same time. However, CWPA has not been investigated with clinical transducers for prostate applications. We report here the development of such a dual-wavelength laser diode-based system and experimental results. We differentiate the acoustic signals resulting from each laser to provide fast and simultaneous spectral image.
In this paper, a novel Delay and Glow (DAG), back projection-based photoacoustic reconstruction, for point source recovery is presented. The projections of the absorber coefficients are delayed and weighed based on the transducer directivity sensitivity and matched with the other projections from other transducer elements leading to glow the point source at the exact location. The DAG was tested with 2-D experimental and simulation studies and compared with conventional delay and sum (DAS) reconstruction method. The proposed algorithm outperforms the DAS in terms of contrast to noise ratio (CNR) and RMS errors.
Iterative image reconstruction algorithms have the potential to reduce the computational time required for photoacoustic tomography (PAT). An iterative deconvolution-based photoacoustic reconstruction with sparsity regularization (iDPARS) is presented which enables us to solve large-scale problems. The method deals with the limited angle of view and the directivity effects associated with clinically relevant photoacoustic tomography imaging with conventional ultrasound transducers. Our Graphics Processing Unit (GPU) implementation is able to reconstruct large 3-D volumes (100×100×100) in less than 10 minutes. The simulation and experimental results demonstrate iDPARS provides better images than DAS in terms of contrast-to-noise ratio and Root-Mean-Square errors.
We define a deconvolution based photoacoustic reconstruction with sparsity regularization (DPARS) algorithm for image restoration from projections. The proposed method is capable of visualizing tissue in the presence of constraints such as the specific directivity of sensors and limited-view Photoacoustic Tomography (PAT). The directivity effect means that our algorithm treats the optically-generated ultrasonic waves based on which direction they arrive at the transducer. Most PA image reconstruction methods assume that sensors have omni-directional response; however, in practice, the sensors show higher sensitivity to the ultrasonic waves coming from one specific direction. In DPARS, the sensitivity of the transducer to incoming waves from different directions are considered. Thus, the DPARS algorithm takes into account the relative location of the absorbers with respect to the transducers, and generates a linear system of equations to solve for the distribution of absorbers. The numerical conditioning and computing times are improved by the use of a sparse Discrete Fourier Transform (DCT) representation of the distribution of absorption coefficients. Our simulation results show that DPARS outperforms the conventional Delay-and-Sum reconstruction method in terms of CNR and RMS errors. Experimental results confirm that DPARS provides images with higher resolution than DAS.
We have integrated photo-acoustic imaging into an automated breast ultrasound scanner (ABUS) with the goal of simultaneously performing ultrasound (US) and multi-spectral photo-acoustic tomography (PAT). This was accomplished with minimal change to the existing automated scanner by coupling laser light into an optical fiber for flexible and robust light delivery. We present preliminary tomography data acquired with this setup, including a simple resolution-testing geometry and a tissue phantom. Integrating PAT into the ABUS such that breast imaging is possible will require illumination from below the transducer dome. To that end, we are moving towards a fiber-based, localized illumination geometry which is fixed relative to the transducer. By illuminating locally (only near the current acquisition slice), this approach reduces overall light exposure at the tissue surface, allowing higher light intensity per acquisition (which translates to higher absorber contrast), while remaining below safe exposure thresholds. We present time-domain simulations of photo-acoustic imaging under non-uniform illumination conditions, and test one potential weighting scheme which can be used to extract absorber locations.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.