Reconstruction using optimal spatially variant kernel for B-mode ultrasound imaging

Desoky, Hesham; Youssef, Abou-Bakr M.; Kadah, Yasser M.

doi:10.1117/12.479902

Cited by 6 publications

(12 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…T is the vector of scatterer amplitudes. This model has been used in B-mode (2-dimensional) [4][5][6][7][8][9], A-mode (1-dimensional) [16,17], and 3-dimensional [18] ultrasound imaging.…”

Section: Model-based Imaging and Regularizationmentioning

confidence: 99%

“…Naturally, real-world inspected objects easily break this assumption and many scatteres may be located off-grid. Many previous studies with model-based algorithms for ultrasound imaging, including but not limited to [4][5][6][7][8][9][10][11], have reported that resolution and contrast are substantially improved in comparison to delay-and-sum (DAS) algorithms when data comes from simulations with scatterers…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Sparse Ultrasound Imaging via Manifold Low-Rank Approximation and Non-Convex Greedy Pursuit

Passarin¹,

Pipa²,

Zibetti³

2018

Preprint

View full text Add to dashboard Cite

Model-based image reconstruction has brought improvements in terms of contrast and spatial resolution to imaging applications such as magnetic resonance imaging and emission computed tomography. However, their use for pulse-echo techniques like ultrasound imaging is limited by the fact that model-based algorithms assume a finite grid of possible locations of scatterers in a medium -- which does not reflect the continuous nature of real world objects and creates a problem known as off-grid deviation. To cope with this problem, we present a method of dictionary expansion and constrained reconstruction that approximates the continuous manifold of all possible scatterer locations within a region of interest. The expanded dictionary is created using a highly coherent sampling of the region of interest, followed by a rank reduction procedure based on a truncated singular value decomposition. We develop a greedy algorithm, based on the Orthogonal Matching Pursuit (OMP), that uses a correlation-based non-convex constraint set that allows for the division of the region of interest into cells of any size. To evaluate the performance of the method, we present results of 2-dimensional ultrasound image reconstructions with simulated data in a nondestructive testing application. Our method succeeds in the reconstructions of sparse images from noisy measurements, providing higher accuracy than previous approaches based on regular discrete models.

show abstract

“…T is the vector of scatterer amplitudes. This model has been used in B-mode (2-dimensional) [4][5][6][7][8][9], A-mode (1-dimensional) [16,17], and 3-dimensional [18] ultrasound imaging.…”

Section: Model-based Imaging and Regularizationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Sparse Ultrasound Imaging via Manifold Low-Rank Approximation and Non-Convex Greedy Pursuit

Passarin¹,

Pipa²,

Zibetti³

2018

Preprint

View full text Add to dashboard Cite

show abstract

“…For mechanically scanned B-scans, examples of this class of methods can be found in [2] in which linear minimum mean squared error (MMSE) was used to estimate the scatter map. A similar method based on singular value decomposition (SVD) regularization is found in [3]. One disadvantage of the methods is that the computation time is typically much larger than for SAFT processing, often limiting their use only to applications without real-time constraints.…”

Section: Introductionmentioning

confidence: 99%

“…One disadvantage of the methods is that the computation time is typically much larger than for SAFT processing, often limiting their use only to applications without real-time constraints. Linear methods as those in [2] and [3] are best adapted to objects containing diffuse scatterers, with a Gaussian distribution modeling the amplitudes of these. In many application, particularly in nondestructive testing (NDT), the images are better described by a collection of a few but relatively strong contribution, for instance, indicating small cracks or material inclusions.…”

Section: Introductionmentioning

confidence: 99%

Sparse Deconvolution of B-Scan Images

Olofsson

Wennerström

2007

IEEE Trans. Ultrason., Ferroelect., Freq. Contr.

View full text Add to dashboard Cite

Abstract-In this paper, a new computationally efficient sparse deconvolution algorithm for the use on B-scan images from objects with relatively few scattering targets is presented. It is based on a linear image formation model that has been used earlier in connection with linear minimum mean squared error (MMSE) two-dimensional (2-D) deconvolution. The MMSE deconvolution results have shown improved resolution compared to synthetic aperture focusing technique (SAFT), but at the cost of increased computation time. The proposed algorithm uses the sparsity of the image, reducing the degrees of freedom in the reconstruction problem, to reduce the computation time and to improve the resolution. The dominating task in the algorithm consists in detecting the set of active scattering targets, which is done by iterating between one up-dating pass that detects new points to include in the set, and a down-dating pass that removes redundant points. In the up-date, a spatiotemporal matched filter is used to isolate potential candidates. A subset of those are chosen using a detection criterion. The amplitudes of the detected scatterers are found by MMSE. The algorithm properties are illustrated using synthetic and real B-scan. The results show excellent resolution enhancement-and noise-suppression capabilities. The involved computation times are analyzed.

show abstract

“…Inverse problem approaches 3 and regularization techniques, such as truncated singular value decomposition (TSVD) , 4 total variation , 5 total least squares 6 and others , 7 have already been used in acoustical imaging and their effectiveness have been explored to a limited extent. It is of special importance to compare the performance of this class of techniques with that which can be obtained using conventional acoustic imaging methods.…”

mentioning

confidence: 99%

A Study of Pulse-Echo Image Formation Using Non-Quadratic Regularization with Speckle-Based Images

2007

View full text Add to dashboard Cite

Abstract:Objective: B-mode imaging is a standard way of presenting results when pulse-echo data acquisition is used. Inverse approaches have been explored to a limited extent. The goals of this study are to explore the feasibility and effectiveness of regularization by: (a) evaluating the quality of the reconstruction of speckle-based images as a function of imaging parameters (i.e., bandwidth of the transducer, f/number, and distance between region of interest (ROI) and focal region), and (b) comparing the reconstructed images with those obtained by conventional imaging techniques (e.g., B-mode and synthetic aperture focusing techniques (SAFT)). Methodology: Twodimensional pulse-echo data from a single-element focused transducer with center frequency of 6 MHz and focal distance of 19 mm were simulated using fractional bandwidths of 33%, 50% and 100%, f/numbers of 1, 2 and 3, and distances between ROI and focus of 0, 2.5 and 5 mm. A minimum scatterer density of 4 scatterers per resolution cell was used. The images were reconstructed using non-quadratic regularization. Because the desired image was known, the normalized mean squared error (MSE) was used as a quantitative indicator of reconstruction performance. The reconstructed images were also evaluated by visual inspection. Results: The best reconstructions were achieved when the ROI was at the focus, the bandwidth was 100% and the f/number was 1, for which the MSE was 25% for a 20-dB signal-to-noise ratio (SNR). The MSE increased as the bandwidth decreased, the f/number increased, and the distance between ROI and focus increased. The bandwidth was the most sensitive parameter, increasing the MSE up to 86% for a SNR of 20 dB. For all cases, parameter selection techniques such as the L-curve and generalized cross-validation (GCV) gave a close

show abstract

Reconstruction using optimal spatially variant kernel for B-mode ultrasound imaging

Cited by 6 publications

References 21 publications

Sparse Ultrasound Imaging via Manifold Low-Rank Approximation and Non-Convex Greedy Pursuit

Sparse Ultrasound Imaging via Manifold Low-Rank Approximation and Non-Convex Greedy Pursuit

Sparse Deconvolution of B-Scan Images

A Study of Pulse-Echo Image Formation Using Non-Quadratic Regularization with Speckle-Based Images

Contact Info

Product

Resources

About