Three-dimensional photoacoustic tomography based on graphics-processing-unit-accelerated finite element method

Enhancing Finite Element-Based Photoacoustic Tomography by Localized Reconstruction Method

Iterative reconstruction algorithm based on finite element (FE) modeling is a powerful approach in photoacoustic tomography (PAT). However, an iterative inverse algorithm using conventional FE meshing of the entire imaging zone is computationally demanding, which hinders this powerful tool in applications where quick image acquisition and/or a large image matrix is needed. To address this challenge, parallel computing techniques are proposed and implemented in the field. Here, we present an alternative approach for 2D PAT, which locoregionally reconstructs the region of interest (ROI) instead of the full imaging zone. Our simulated and phantom experimental results demonstrate that this ROI reconstruction algorithm can produce almost the same image quality as the conventional full zone-based reconstruction algorithm; however, the computation time can be significantly reduced without any additional hardware cost by more than two orders of magnitude (100-fold). This algorithm is further applied and validated in an in vivo study. The major vessel structures in a rat’s brain can be imaged clearly using our ROI-based approach, coupled with a mesh of 11,801 nodes. This novel algorithm can also be parallelized using MPI or GPU acceleration techniques to further enhance the reconstruction performance of FE-based PAT.

Model-Based Reconstruction of Large Three-Dimensional Optoacoustic Datasets

Iterative model-based algorithms are known to enable more accurate and quantitative optoacoustic (photoacoustic) tomographic reconstructions than standard back-projection methods. However, three-dimensional (3D) model-based inversion is often hampered by high computational complexity and memory overhead. Parallel implementations on a graphics processing unit (GPU) have been shown to efficiently reduce the memory requirements by on-the-fly calculation of the actions of the optoacoustic model matrix, but the high complexity still makes these approaches impractical for large 3D optoacoustic datasets. Herein, we show that the computational complexity of 3D model-based iterative inversion can be significantly reduced by splitting the model matrix into two parts: one maximally sparse matrix containing only one entry per voxel-transducer pair and a second matrix corresponding to cyclic convolution. We further suggest reconstructing the images by multiplying the transpose of the model matrix calculated in this manner with the acquired signals, which is equivalent to using a very large regularization parameter in the iterative inversion method. The performance of these two approaches is compared to that of standard back-projection and a recently introduced GPU-based model-based method using datasets from in vivo experiments. The reconstruction time was accelerated by approximately an order of magnitude with the new iterative method, while multiplication with the transpose of the matrix is shown to be as fast as standard back-projection.

Graphics processing unit accelerating compressed sensing photoacoustic computed tomography with total variation

Photoacoustic computed tomography with compressed sensing (CS-PACT) is a commonly used imaging strategy for sparse-sampling PACT. However, it is very time-consuming because of the iterative process involved in the image reconstruction. In this paper, we present a graphics processing unit (GPU)-based parallel computation framework for total-variation-based CS-PACT and adapted into a custom-made PACT system. Specifically, five compute-intensive operators are extracted from the iteration algorithm and are redesigned for parallel performance on a GPU. We achieved an image reconstruction speed 24-31 times faster than the CPU performance. We performed in vivo experiments on human hands to verify the feasibility of our developed method.

Parallel Computing for Quantitative Blood Flow Imaging in Photoacoustic Microscopy

Photoacoustic microscopy (PAM) is an emerging biomedical imaging technology capable of quantitative measurement of the microvascular blood flow by correlation analysis. However, the computational cost is high, limiting its applications. Here, we report a parallel computation design based on graphics processing unit (GPU) for high-speed quantification of blood flow in PAM. Two strategies were utilized to improve the computational efficiency. First, the correlation method in the algorithm was optimized to avoid redundant computation and a parallel computing structure was designed. Second, the parallel design was realized on GPU and optimized by maximizing the utilization of computing resource in GPU. The detailed timings and speedup for each calculation step were given and the MATLAB and C/C++ code versions based on CPU were presented as a comparison. Full performance test shows that a stable speedup of ~80-fold could be achieved with the same calculation accuracy and the computation time could be reduced from minutes to just several seconds with the imaging size ranging from 1 × 1 mm² to 2 × 2 mm². Our design accelerates PAM-based blood flow measurement and paves the way for real-time PAM imaging and processing by significantly improving the computational efficiency.

Finite-element reconstruction of 2D circular scanning photoacoustic tomography with detectors in far-field condition

The finite-element method (FEM) has been a powerful numerical tool for the reconstruction of 2D circular scanning-based photoacoustic tomography (PAT) for its unrivaled ability to accommodate complex boundary conditions, quantitatively reconstruct different physical parameters, and enable low sampling frequency and fewer detector numbers. To reduce the computation cost, a smaller image domain is commonly used instead of the region surrounded by the transducer scanning trace. Then, the pressure data used for the reconstruction that is defined on the boundary of the image domain is usually obtained by directly time delaying the actual measured data. In this case, distortions will be aroused for targets that are away from the rotation center. In this work, we put forward a new data preprocessing method to overcome this problem with a virtual detector concept, in which the measured data for the virtual point detectors on the boundary of the reconstruction domain are generated by a summation of the signals from nearby true detectors. The complete removal of the distortions using our proposed algorithm was proven with experimental reconstruction results.

Real-time GPU-accelerated processing and volumetric display for wide-field laser-scanning optical-resolution photoacoustic microscopy

Fast signal processing and real-time displays are essential for practical imaging modality in various fields of applications. However, the imaging speed in optical-resolution photoacoustic microscopy (OR-PAM), in particular, depends on factors such as the pulse repetition rate of the laser, scanning method, field of view (FOV), and signal processing time. In the past, efforts to increase acquisition speed either focused on developing new scanning methods or using lasers with higher pulse repetition rates. However, high-speed signal processing is also important for real-time volumetric display in OR-PAM. In this study, we carried out parallel signal processing using a graphics processing unit (GPU) to enable fast signal processing and wide-field real-time displays in laser-scanning OR-PAM. The average total GPU processing time for a B-mode PAM image was approximately 1.35 ms at a display speed of 480 fps when the data samples were acquired with 736 (axial) × 500 (lateral) points/B-mode-frame at a pulse repetition rate of 300 kHz. In addition, we successfully displayed maximum amplitude projection images of a mouse's ear as volumetric images with an FOV of 3 mm × 3 mm (500 × 500 pixels) at 1.02 s, corresponding to 0.98 fps.

Semi-quantitative Multispectral Optoacoustic Tomography (MSOT) for volumetric PK imaging of gastric emptying

A common side effect of medication is gastrointestinal intolerance. Symptoms can include reduced appetite, diarrhea, constipation, GI inflammation, nausea and vomiting. Such effects often have a dramatic impact on compliance with a treatment regimen. Therefore, characterization of GI tolerance is an important step when establishing a novel therapeutic approach. In this study, Multispectral Optoacoustic Tomography (MSOT) is used to monitor gastrointestinal motility by in vivo whole body imaging in mice. MSOT combines high spatial and temporal resolution based on ultrasound detection with strong optical contrast in the near infrared. Animals were given Indocyanine Green (ICG) by oral gavage and imaged by MSOT to observe the fate of ICG in the gastrointestinal tract. Exponential decay of ICG signal was observed in the stomach in good correlation with ex vivo validation. We discuss how kinetic imaging in MSOT allows visualization of parameters unavailable to other imaging methods, both in 2D and 3D.