Photoacoustic imaging method based on arc-direction compressed sensing and multi-angle observation

Abstract

In photoacoustic imaging (PAI), the photoacoustic (PA) signal can be observed only from limit-view angles due to some structure limitations. As a result, data incompleteness artifacts appear and some image details lose. An arc-direction mask in PA data acquisition and arc-direction compressed sensing (CS) reconstruction algorithm are proposed instead of the conventional rectangle CS methods for PAI. The proposed method can effectively realize the compression of the PA data along the arc line and exactly recover the PA images from multi-angle observation. Simulation results demonstrate that it has the potential of application in high-resolution PAI for obtaining highly resolution and artifact-free PA images.

Compressed-sensing photoacoustic computed tomography in vivo with partially known support

“…Fortunately, most medical images are sparse in a certain domain by finding an appropriate sparse transform : = ψ x ψθ , where θ is the original image, and x is the transformed one. It has been proven that photoacoustic images can be transformed into sparse domains by a variety of transforms, such as the numerical derivative (ND), the wavelet transform, etc [16][17][18][19].…”

“…Recently, explorations aiming to leverage the advantages of CS for PACT have also been reported. For example, J. Provost et al have conducted numerical and phantom studies on CS-based PACT [16]; D. Liang et al have numerically demonstrated CS-based photoacoustic imaging under random optical illumination [17]; Z. Guo et al have reported CS-based photoacoustic reconstruction in time domain with images acquired in vivo [18]; M. Sun et al have developed an arc-direction compressed-sensing PACT algorithm with numerical phantoms [19]. All these studies have shown that CS-based reconstruction techniques can recover photoacoustic signals sampled at a (temporal and/or spatial) rate lower than that required by the Nyquist sampling theory.…”

Sparsity‐based acoustic inversion in cross‐sectional multiscale optoacoustic imaging

“…(3) is based on the common assumption that the reconstructed image may be sparsely represented in an alternative basis. [24][25][26][27][28][29] To perform the inversion, the following procedure is applied. First, the image to be recovered u is transferred into a basis in which it is expected to be sparse: x = Φu, where x is the representation of u in the new domain and Φ is the transform matrix.…”

“…More recent regularization approaches are based on either penalizing the total variation (TV) of the image [20][21][22][23] or enforcing sparsity on its representation in an alternative basis, e.g., a wavelet basis. [24][25][26][27][28][29] Both of these regularization approaches have been shown to produce sharper, cleaner images than energy-minimization regularization in several imaging scenario, and their combination has been used in image denoising. 30 Nonetheless, demonstration on experimental data has been limited to only tissue-mimicking phantoms 29 or experiments in mice that focused on imaging vascularization.…”

Compressed Sensing Photoacoustic Imaging Based on Fast Alternating Direction Algorithm

“…The issue of artifacts and loss of resolution in limited-view imaging can be addressed by using random optical illuminations for fast data acquisition via the SPGL1 algorithm [15, 16]. Sun et al have developed an arc-direction compressed-sensing PAT algorithm with numerical phantoms [17]. Both phantom and in vivo results showed that the CS method can effectively reduce undersampling artifacts via the nonlinear conjugate gradient descent algorithm [18, 19].…”