Je Lindop

A quality-guided displacement tracking algorithm for ultrasonic elasticity imaging.

Displacement estimation is a key step in the evaluation of tissue elasticity by quasistatic strain imaging. An efficient approach may incorporate a tracking strategy whereby each estimate is initially obtained from its neighbours’ displacements and then refined through a localized search. This increases the accuracy and reduces the computational expense compared with exhaustive search. However, simple tracking strategies fail when the target displacement map exhibits complex structure. For example, there may be discontinuities and regions of indeterminate displacement caused by decorrelation between the pre- and post-deformation radio frequency (RF) echo signals. This paper introduces a novel displacement tracking algorithm, with a search strategy guided by a data quality indicator. Comparisons with existing methods show that the proposed algorithm is more robust when the displacement distribution is challenging.

Phase-based ultrasonic deformation estimation

Deformation estimation is the foundation of emerging techniques for imaging the mechanical properties of soft tissues. We present theoretical analysis and experimental results from an investigation of phase-based ultrasonic deformation estimators. Numerous phase-based algorithm variants were tested quantitatively on simulated R.F data from uniform scatterer fields, subject to a range of uniform strain deformations. Particular attention is paid to a new algorithm, weighted phase separation, the performance of which is demonstrated in application to in vivo freehand strain imaging. Good results support the theory that underlies the new algorithm, and more generally highlight the factors that should be considered in the design of high-performance deformation estimators for practical applications. For context, note that this represents progress with an algorithm class that is suitable for real-time applications, yet has already been shown quantitatively to offer greater accuracy over a wide range of scanning conditions than adaptive companding methods based on correlation coefficient or sum of absolute differences.

Estimation of Displacement Location for Enhanced Strain Imaging

Ultrasonic strain imaging usually begins with displacement estimates computed using finite-length sections of RF ultrasound signals. Amplitude variations in the ultrasound are known to perturb the location at which the displacement estimate is valid. If this goes uncorrected, it is a significant source of estimation noise, which is amplified when displacement fields are converted into strain images. We present a study of this effect based on theoretical analysis and practical experiments. A correction method based on the analysis is tested on phase zero and correlation coefficient strain imaging, and compared to the amplitude compression techniques of earlier studies. We also test adaptive strain estimation to provide a benchmark, but the performance of our new method matches or surpasses this benchmark under normal scanning conditions. Furthermore, the new correction is suitable for real time applications owing to its extreme computational simplicity.

An intelligent interface for freehand strain imaging.

We present a new, intelligent interface for freehand strain imaging, which has been designed to support clinical trials investigating the potential of ultrasonic strain imaging for diagnostic purposes across a broad range of target pathologies. The aim with this interface is to make scanning easier and to help clinicians learn the necessary scanning technique quickly, by providing real time feedback indicating the quality of the strain data as they are produced. The methods require a pixel-level indicator of estimation precision, which can be calculated in-line with strain estimation. This is exploited in novel approaches to normalisation, persistence and display. The effect of each component is indicated in the results with examples from in vitro and in vivo scanning. As well as providing real-time feedback, the images are easier to interpret because data at unacceptably low signal-to-noise ratios do not reach the display. Additionally, the level of noise in the displayed images is actually reduced compared with other methods that use the same strain estimates with the same level of persistence. The interface also considerably reduces the difficulty in producing volumes of strain data from freehand three-dimensional scans.

Dynamic Resolution Selection in Ultrasonic Strain Imaging

Ultrasonic strain imaging promises to be a valuable tool in medical diagnostics. Reliability and ease-of-use have become important considerations. These depend on selection of appropriate imaging parameters. Two tasks are undertaken here. The tradeoff between resolution and estimation precision is examined closely to establish models for the relationships with imaging parameters and data properties. These models are then applied in a system that automatically sets the imaging parameters responsive to the data quality and the required estimation precision, so as to produce more meaningful images under varying scan conditions. The new system is applied to simulation, in vitro and in vivo data for validation. It reduces the complexity of the sonographers role in strain imaging, and produces images of reliable quality even when the level of signal decorrelation varies throughout the ultrasound data.

The general properties including accuracy and resolution of linear filtering methods for strain estimation

The vast majority of strain imaging systems applies linear filtering to estimate strain from displacement data. Methods such as piecewise-linear least squares regression and staggered strain estimation have come to be widely known and applied, but the properties of these estimators have rarely (or never) been compared quantitatively. Given their tractable properties, careful analysis of linear filters allows us to make numerous observations that are simple, yet valuable. We consider accuracy and resolving power, which raises the question of whether any particular filter offers the best possible accuracy at a given resolution. Our surprising results provide insight at two levels: They highlight general considerations affecting the type of filter that is appropriate for practical applications, and indicate promising avenues for further research.

Uniform precision ultrasound strain imaging

Ultrasound strain imaging is becoming increasingly popular as a way to measure stiffness variation in soft tissue. Almost all techniques involve the estimation of a field of relative displacements between measurements of tissue undergoing different deformations. These estimates are often high resolution, but some form of smoothing is required to increase the precision, either by direct filtering or as part of the gradient estimation process. Such methods generate uniform resolution images, but strain quality typically varies considerably within each image, hence a trade-off is necessary between increasing precision in the low-quality regions and reducing resolution in the high-quality regions. We introduce a smoothing technique, developed from the nonparametric regression literature, which can avoid this trade-off by generating uniform precision images. In such an image, high resolution is retained in areas of high strain quality but sacrificed for the sake of increased precision in low-quality areas. We contrast the algorithm with other methods on simulated, phantom, and clinical data, for both 2-D and 3-D strain imaging. We also show how the technique can be efficiently implemented at real-time rates with realistic parameters on modest hardware. Uniform precision nonparametric regression promises to be a useful tool in ultrasound strain imaging.

Stable, intelligible ultrasonic strain imaging.

Background: Freehand quasistatic strain imaging can reveal qualitative information about tissue stiffness with good spatial accuracy. Clinical trials, however, repeatedly cite instability and variable signal-to-noise ratio as significant drawbacks. Methods: This study investigates three post-processing strategies for quasistatic strain imaging. Normalization divides the strain by an estimate of the stress field, the intention being to reduce sensitivity to variable applied stress. Persistence aims to improve the signal-to-noise ratio by time-averaging multiple frames. The persistence scheme presented in this article operates at the pixel level, weighting each frames contribution by an estimate of the strain precision. Precision-based display presents the clinician with an image in which regions of indeterminate strain are obscured behind a colour wash. This is achieved using estimates of strain precision that are faithfully propagated through the various stages of signal processing. Results and discussion: The post-processing strategy is evaluated qualitatively on scans of a breast biopsy phantom and in vivo head and neck examinations. Strain images processed in this manner are observed to benefit from improved stability and signal-to-noise ratio. There are, however, limitations. In unusual though conceivable circumstances, the normalization procedure might suppress genuine stiffness variations evident in the unprocessed strain images. In different circumstances, the raw strain images might fail to capture significant stiffness variations, a situation that no amount of post-processing can improve. Conclusion: The clinical utility of freehand quasistatic strain imaging can be improved by normalization, precision-weighted pixel-level persistence and precision-based display. The resulting images are stable and generally exhibit a better signal-to-noise ratio than any of the original, unprocessed strain images.

Deconvolution and elastography based on three-dimensional ultrasound

This paper is in two parts and addresses two ways of getting more information out of the RF signal from a three-dimensional (3D) mechanically-swept medical ultrasound scanner. The first topic is the use of non-blind deconvolution to improve the clarity of the data, particularly in the direction perpendicular to the individual B-scans. The second topic is strain imaging. We present a robust and efficient approach to the estimation and display of axial strain information. For deconvolution, we calculate an estimate of the point-spread function at each depth in the image using Field II. This is used as part of an Expectation Maximisation (EM) framework in which the ultrasound scatterer field is modelled as the product of (a) a piecewise smooth function and (b) a fine-grain varying function. In the E step, a Wiener filter is used to estimate the scatterer field based on an assumed piecewise smooth component. In the M step, wavelet de-noising is used to estimate the piecewise smooth component from the scatterer field. For strain imaging, we use a quasi-static approach with efficient phase-based algorithms. Our contributions lie in robust and efficient 3D displacement tracking, point-wise quality-weighted averaging, and a stable display that shows not only strain but also an indication of the quality of the data at each point in the image. This enables clinicians to see where the strain estimate is meaningful and where it is mostly noise. For deconvolution, we present in-vivo images and simulations with quantitative performance measures. With the blurred 3D data taken as 0 dB, we get an improvement in signal to noise ratio of 4.6 dB with a Wiener filter alone, 4.36 dB with the ForWaRD algorithm and 5.18 dB with our EM algorithm. For strain imaging we show images based on 2D and 3D data and describe how full 3D analysis can be performed in about 20 seconds on a typical computer. We will also present initial results of our clinical study to explore the applications of our system in our local hospital.

Near-Real-Time 3D Ultrasonic Strain Imaging

This paper describes a near-real-time system for acquiring and displaying 3D ultrasonic strain images using a mechanical sector transducer. For improved image quality and robustness, all signal processing is fully 3D, including 3D data windows, 3D least-squares fitting for the displacement-to-strain calculation, 3D strain normalization and full displacement tracking in the axial, lateral and elevational directions. Notwithstanding this thorough signal processing, 3D strain volumes are typically available for inspection within 20 seconds of performing the scan, with no need for special hardware. The speed is achieved by iterative phase-zero displacement tracking in the axial direction and novel methods for tracking in the lateral and elevational directions. Since the displacement tracking does not rely on the common (but brittle) zero-displacement assumption at the transducer face, high quality strain images are obtained reliably. The paper includes examples of in vitro strain images with full details of the acquisition and processing times.