Ultrasonic imaging using a computed point spread function

Rangarajan, Ramsharan; Krishnamurthy, C. V.; Balasubramaniam, Krishnan

doi:10.1109/tuffc.2008.663

Cited by 14 publications

(13 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Its complexity would thus be greater than the square of the number of pixels in the image, limiting its applicability to medium sized images. Using the matrix-free expression in (2), operator H performs m k n k m t n t multiplications and has negligible memory requirements. Therefore, in ultrasound imaging, the matrix-free representation is not only vastly superior to its matrix counterpart (because the kernel is much smaller than the image), but has the same computational complexity as spatially invariant convolution (excluding the unrealistic circulant boundary case).…”

Section: B Axially Varying Convolutionmentioning

confidence: 99%

An Axially Variant Kernel Imaging Model Applied to Ultrasound Image Reconstruction

Florea

Basarab

Kouamé

et al. 2018

IEEE Signal Process. Lett.

View full text Add to dashboard Cite

Existing ultrasound deconvolution approaches unrealistically assume, primarily for computational reasons, that the convolution model relies on a spatially invariant kernel and circulant boundary conditions. We discard both restrictions and introduce an image formation model applicable to ultrasound imaging and deconvolution based on an axially varying kernel, that accounts for arbitrary boundary conditions. Our model has the same computational complexity as the one employing spatially invariant convolution and has negligible memory requirements. To accommodate state-of-the-art deconvolution approaches when applied to a variety of inverse problem formulations, we also provide an equally efficient adjoint expression of our model. Simulation results confirm the tractability of our model for the deconvolution of large images. Moreover, the quality of reconstruction using our model is superior to that obtained using spatially invariant convolution.

show abstract

Section: B Axially Varying Convolutionmentioning

confidence: 99%

An Axially Variant Kernel Imaging Model Applied to Ultrasound Image Reconstruction

Florea

Basarab

Kouamé

et al. 2018

IEEE Signal Process. Lett.

View full text Add to dashboard Cite

show abstract

“…Many authors thus discuss deconvolution in conjunction with PSF assessment [1]- [3]. Two approaches to account for PSF blurring can be distinguished: deterministic [1], [2], [4] and blind deconvolution [3], [5].…”

Section: Background and Motivationmentioning

confidence: 99%

“…To avoid too complex an optimization, the PSF is usually assumed spatially-invariant across the imaging domain. In non-blind deconvolution [4], [6], a deterministic model of the PSF is obtained by means of simulation, such as Field II [7] or numerical approximation [1]. Again, the convenience of a spatiallyinvariant PSF avoids time-consuming repeated simulation.…”

mentioning

confidence: 99%

On an analytical, spatially-varying, point-spread-function

Roquette

Simeoni

Hurley

et al. 2017

2017 IEEE International Ultrasonics Symposium (IUS)

View full text Add to dashboard Cite

Abstract-The point spread function (PSF), namely the response of an ultrasound system to a point source, is a powerful measure of the quality of an imaging system. The lack of an analytical formulation inhibits many applications ranging from apodization optimization, array-design, and deconvolution algorithms. We propose to fill this gap through a general PSF derivation that is flexible with respect to the type of transmission (synthetic aperture, plane-wave, diverging-wave etc.), while faithfully capturing the spatially-variant blurring of the Tissue Reflectivity Function as caused by Delay-And-Sum reconstruction. We validate the derived PSF against simulation using Field II, and show that accounting for PSF spatial-variability in sparsebased deconvolution improves reconstruction.Index Terms-Point-Spread-Function, deconvolution, image enhancement, ultrasonic image simulation, apodization design. I. BACKGROUND AND MOTIVATIONIn ultrasound (US) imaging, the finite bandwidth and aperture of transducer elements limits image resolution. This limitation on the resolving capability of the tissue reflectivity function (TRF) can be modelled explicitly by re-casting the radio-frequency (RF) image as the results of an operator between the point-spread function (PSF) and the TRF. In this sense, the PSF, defined as the response of the imaging method in presence of a single scatterer, contains the blurring due to the instrument, and is a powerful tool in assessing the equipment performance in terms of imaging quality.Deconvolution methods use RF images to retrieve the TRF by accounting for the effect of PSF and prior knowledge of the image type. Many authors thus discuss deconvolution in conjunction with PSF assessment [1]- [3]. Two approaches to account for PSF blurring can be distinguished: deterministic Most deconvolution methods, blind or not, assume a spatially-invariant PSF model, primarily for computational purposes. In the blind-deconvolution context, [3], [5] argue that tissue-dependent attenuation and dispersive effects require the PSF to be estimated during the TRF computation. To avoid too complex an optimization, the PSF is usually assumed spatially-invariant across the imaging domain. In non-blind deconvolution [4], [6], a deterministic model of the PSF is obtained by means of simulation, such as Field II [7] or numerical approximation [1]. Again, the convenience of a spatiallyinvariant PSF avoids time-consuming repeated simulation.In both cases, a spatial invariance assumption on the PSF hence leads to important computational gain, since the operator linking the TRF and RF image becomes a convolution, which can efficiently be implemented in the Fourier domain. This assumption can however have profound consequences on the quality of the recovered TRF images: in practice, the PSF can indeed vary quite dramatically across the imaging domain, leading to non-uniform recovery performances.A few attempts have been made to account for spatiallyvarying PSFs [3]. However, they usually come at the cost of some simplifying...

show abstract

“…A possible solution to the limited depth range could be the use of double-sided inspection [27], but this is on the expense of a more involved experimental procedure. In [28] and [29], the point-spread function of ultrasonic pulse-echo imaging is computed and used in different deconvolution schemes to enhance spatial resolution. In [30], an axially varying kernel forward convolution model is applied for the deconvolution of ultrasonic imaging yielding enhanced results over fixed-kernel methods.…”

Section: Introductionmentioning

confidence: 99%