Full-wave attenuation reconstruction in the time domain for ultrasound computed tomography (original) (raw)
Related papers
The Journal of the Acoustical Society of America, 2017
Ultrasound computed tomography (USCT) is a non-invasive imaging technique that provides information about the acoustic properties of soft tissues in the body, such as the speed of sound (SS) and acoustic attenuation (AA). Knowledge of these properties can improve the discrimination between benign and malignant masses, especially in breast cancer studies. Full wave inversion (FWI) methods for image reconstruction in USCT provide the best image quality compared to more approximate methods. Using FWI, the SS is usually recovered in the time domain, and the AA is usually recovered in the frequency domain. Nevertheless, as both properties can be obtained from the same data, it is desirable to have a common framework to reconstruct both distributions. In this work, an algorithm is proposed to reconstruct both the SS and AA distributions using a time domain FWI methodology based on the fractional Laplacian wave equation, an adjoint field formulation, and a gradient-descent method. The optimization code employs a Compute Unified Device Architecture version of the software k-Wave, which provides high computational efficiency. The performance of the method was evaluated using simulated noisy data from numerical breast phantoms. Errors were less than 0.5% in the recovered SS and 10% in the AA. V
Improved misfit function for attenuation and speed reconstruction in ultrasound computed tomography
Medical Imaging 2017: Ultrasonic Imaging and Tomography, 2017
The reconstruction of acoustic attenuation maps for transmission Ultrasound Computed Tomography (USCT) based on the standard least-squares full wave inversion method requires the accurate knowledge of the sound speed map in the region under study. Any deviation in the reconstructed speed maps creates a very significant bias in the attenuation map, as the standard least-squares misfit function is more sensitive to time misalignments than to amplitude differences of the signals. In this work, we propose a generalized misfit function which includes an additional term that accounts for the amplitude differences between the measured and the estimated signals. The functional gradients used to minimize the proposed misfit function were obtained using an adjoint field formulation and the fractional Laplacian wave equation. The forward and backward wave propagation was obtained with the parallelized GPU version of the software k-Wave and the optimization was performed with a line search method. A numerical phantom simulating breast tissue and synthetic noisy data were used to test the performance of the proposed misfit function. The attenuation was reconstructed based on a converged speed map. An edge-preserving regularization method based on total variation was also implemented. To quantify the quality of the results, the mean values and their standard deviations in several regions of interest were analyzed and compared to the reference values. The proposed generalized misfit function decreases considerably the bias in the attenuation map caused by the deviations in the speed map in all the regions of interest analyzed.
Towards Attenuation Imaging with Computed Ultrasound Tomography in Echo Mode (CUTE)
This work presents a novel attenuation imaging technique for pulse-echo ultrasound systems. In contrast to state-of-the-art techniques, we formulate the reconstruction in two dimensions relying on tissue insonifications with different steering angles. By beamforming backscattered echoes recorded by the transducer, we measure at each location the changes in the amplitudes of detected echoes with different transmissions and relate them to local tissue attenuation. This relationship assumes ultrasound waves propagate in straight paths; thus, we linearize the forward problem to provide suitable time-to-solutions for clinical practice. The presented technique is the natural extension of computed tomography in echo mode (CUTE), initially developed for tissue speed-of-sound quantification. The performance of our method is demonstrated in numerical examples with data computed using the k-Wave numerical solver for wave-propagation simulations. These examples consider tissue-mimicking media w...
Ultrasound Image Reconstruction Using Nesterov's Accelerated Gradient
Ultrasound Image Reconstruction Using Nesterov's Accelerated Gradient, 2018
The aim of this thesis is to conduct research on Nesterov’s accelerated gradient method for the reconstruction of speed of sound as well as attenuation profiles in ultrasound computed tomography. Firstly, ultrasound (acoustic) wave propagation based on paraxial approximation has been performed as the forward model. For iterative reconstruction, exact measurements have been simulated from the forward model and then compared with the estimated measurements which are updated for each iteration based on the reconstructed profiles. This process is known as an inverse problem, which is tackled via minimizing the deviation between exact measurements and estimated measurements, i.e. via solving a nonlinear least-squares problem. To minimize this deviation, Nesterov’s accelerated gradient method has been performed and compared with other optimization algorithms including gradient descent and Gauss-Newton conjugate gradient. Also, two line search (LS) methods have been used to choose the step size for each iteration since finding proper step size is crucial for the convergence of such optimization algorithms. One line search method is backtracking and the other is based on zoom functions. The Wolfe conditions and strong Wolfe conditions have been adopted as the termination condition for line search. In total, seven methods of different combinations of the above algorithms have been tested. These methods are Gauss-Newton conjugate gradient, gradient descent with fixed step size, Nesterov’s accelerated gradient with fixed step size, gradient descent with backtracking LS step size under Wolfe conditions, Nesterov’s accelerated gradient with backtracking LS step size under Wolfe conditions, gradient descent with zoom LS step size under strong Wolfe conditions, Nesterov’s accelerated gradient with zoom LS step size under strong Wolfe conditions. Among the seven methods, Nesterov’s accelerated gradient with LS step size has the fastest convergence rate (iteration number) compared to other methods. However, due to the increased computational complexity of LS for each iteration, it requires extra computational time. On the other side, Nesterov’s accelerated gradient with a fixed step size is the fastest method among all the tested methods regarding computational time. We conclude that Nesterov’s accelerated gradient is a promising algorithm for the image reconstruction in ultrasound transmission tomography, due to its relatively cheap computation per iteration as compared with the state-of-the-art Gauss-Newton conjugate gradient method.
Acoustic attenuation imaging of tissue bulk properties with a priori information
The Journal of the Acoustical Society of America, 2016
Attenuation of ultrasound waves traversing a medium is not only a result of absorption and scattering within a given tissue, but also of coherent scattering, including diffraction, refraction, and reflection of the acoustic wave at tissue boundaries. This leads to edge enhancement and other artifacts in most reconstruction algorithms, other than 3D wave migration with currently impractical, implementations. The presented approach accounts for energy loss at tissue boundaries by normalizing data based on variable sound speed, and potential density, of the medium using a k-space wave solver. Coupled with a priori knowledge of major sound speed distributions, physical attenuation values within broad ranges, and the assumption of homogeneity within segmented regions, an attenuation image representative of region bulk properties is constructed by solving a penalized weighted least squares optimization problem. This is in contradistinction to absorption or to conventional attenuation coef...
High resolution 3D ultrasonic breast imaging by time-domain full waveform inversion
Inverse Problems, 2021
Ultrasound tomography (UST) scanners allow quantitative images of the human breast’s acoustic properties to be derived with potential applications in screening, diagnosis and therapy planning. Time domain full waveform inversion (TD-FWI) is a promising UST image formation technique that fits the parameter fields of a wave physics model by gradient-based optimization. For high resolution 3D UST, it holds three key challenges: firstly, its central building block, the computation of the gradient for a single US measurement, has a restrictively large memory footprint. Secondly, this building block needs to be computed for each of the 103–104 measurements, resulting in a massive parallel computation usually performed on large computational clusters for days. Lastly, the structure of the underlying optimization problem may result in slow progression of the solver and convergence to a local minimum. In this work, we design and evaluate a comprehensive computational strategy to overcome the...
2014 IEEE International Ultrasonics Symposium, 2014
Accurate estimation of the local acoustic attenuation based on the backscatter signal has several applications, e.g. ultrasound tissue characterization. Most of the existing techniques determine the attenuation coefficient of the tissue directly from the spectrum of the backscattered signal. In these approaches other effects, such as diffraction, that may influence the attenuation estimation should be corrected for. This correction may be impractical in vivo. In the present study the simulation of ultrasound wave propagation was used for the estimation of the attenuation characteristics. Indeed, the local attenuation coefficient was estimated by iteratively solving the forward wave propagation problem and matching the synthetic backscattered signal to the measured one. The proposed methodology was experimentally validated using tissuemimicking phantoms with different attenuation characteristics showing promising results.
Regularized Image Reconstruction for Ultrasound Attenuation Transmission Tomography
2008
The paper is focused on ultrasonic transmission tomography as a potential medical imaging modality, namely for breast cancer diagnosis. Ultrasound attenuation coefficient is one of the tissue parameters which are related to the pathological tissue state. A technique to reconstruct images of attenuation distribution is presented. Furthermore, an alternative to the commonly used filtered backprojection or algebraic reconstruction techniques is proposed. It is based on regularization of the image reconstruction problem which imposes smoothness in the resulting images while preserving edges. The approach is analyzed on synthetic data sets. The results show that it stabilizes the image restoration by compensating for main sources of estimation errors in this imaging modality.
A Conjugate Gradient-Neural Network Technique for Ultrasound Inverse Imaging
Journal of Computational Acoustics, 2002
In this paper, a new technique for solving the two-dimensional inverse scattering problem for ultrasound inverse imaging is presented. Reconstruction of a two-dimensional object is accomplished using an iterative algorithm which combines the conjugate gradient (CG) method and a neural network (NN) approach. The neural network technique is used to exploit knowledge of the statistical characteristics of the object to enhance the performance of the conjugate gradient method. The results for simulations show that the CGNN algorithm is more accurate than the CG method and, in addition, convergence occurs more rapidly. For the CGNN algorithm, approximately 50% fewer iterations are needed to obtain the inverse solution for a signal-to-noise ratio (SNR) of 50 dB. For a smaller SNR of 35 dB, the CGNN method is not as accurate, but it still gives reasonable results.
Ultrasonic Computed Tomography
Bone Quantitative Ultrasound, 2010
Ultrasonic Computed Tomography (UCT) is a full digital imaging technique, which consists in numerically solving the inverse scattering problem associated to the forward scattering problem describing the interaction of ultrasonic waves with inhomogeneous media. For weakly inhomogeneous media such as soft tissues, various approximations of the solution of the forward problem (straight ray approximation, Born approximation…), leading to easy-to-implement approximations of the inverse scattering problem (back-projection or backpropagation algorithms) can be used. In the case of highly heterogeneous media such as bone surrounded by soft tissues, such approximations are no more valid. We present here two non-linear inversion schemes based on high-order approximations. These methods are conceived like the prolongation of the methods implemented in the weakly inhomogeneous case for soft tissues. The results show the feasibility of this UCT approach to bones and its potential to perform measurements in vivo.