Model-Based Reconstruction Integrated With Fluence Compensation for Photoacoustic Tomography

Abstract

Photoacoustic (PA) tomography (PAT) is a rapidly developing imaging modality that can provide high contrast and spatial-resolution images of light-absorption distribution in tissue. However, reconstruction of the absorption distribution is affected by nonuniform light fluence. This paper introduces a reconstruction method for reducing amplification of noise and artifacts in low-fluence regions. In this method, fluence compensation is integrated into model-based reconstruction, and the absorption distribution is iteratively updated. At each iteration, we calculate the residual between detected PA signals and the signals computed by a forward model using the initial pressure, which is the product of estimated voxel value and light fluence. By minimizing the residual, the reconstructed values converge to the true absorption distribution. In addition, we developed a matrix compression method for reducing memory requirements and accelerating reconstruction speed. The results of simulation and phantom experiments indicate that the proposed method provides a better contrast-to-noise ratio (CNR) in low-fluence regions. We expect that the capability of increasing imaging depth will broaden the clinical applications of PAT.

Accelerating image reconstruction in three-dimensional optoacoustic tomography on graphics processing units

“…22 Though we described our parallelization strategies for the projection/backprojection operators that utilized two discrete-to-discrete imaging models, these strategies can also be applied to other D-D imaging models. 9,11,20,59 Therefore, the proposed algorithms will facilitate the further investigation and application of advanced image reconstruction algorithms in 3D OAT.…”

“…33 In OAT, however, only a few works on utilization of GPUs to accelerate image reconstruction have been reported. 20,34 For example, the k-wave toolbox employs the NVIDIA CUDA Fast Fourier Transform library (cuFFT) to accelerate the computation of 3D FFT. 34 Also a GPU-based sparse matrix-vector multiplication strategy has been applied to 3D OAT image reconstruction for the case that the system matrix is sparse and can be stored in memory.…”

“…34 Also a GPU-based sparse matrix-vector multiplication strategy has been applied to 3D OAT image reconstruction for the case that the system matrix is sparse and can be stored in memory. 20 However, there remains an important need to develop efficient implementations of OAT reconstruction algorithms for general applications in which the system matrix is too large to be stored.…”

“…To overcome these limitations, iterative image reconstruction algorithms have been proposed. [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24] Although the optoacoustic wave intrinsically propagates in 3D space, when applying to experimental data, most studies have employed twodimensional (2D) imaging models by making certain assumptions on the transducer responses and/or the object structures. 10,13,15,16,19,25 An important reason is because the computation required for 3D OAT image reconstruction is excessively burdensome.…”

Acoustic Inversion in Optoacoustic Tomography: A Review

“…One of the most notable hardware improvements has been the application of GPUs in model-based inversion, which was implemented for the computationally demanding TV regularization [118]. Memory requirements could be reduced by the matrix compression method [41] or on-the-fly calculation of matrix elements [31]. Fast inversion algorithms include the numerically efficient LSQR as well as the pseudo-inverse technique.…”