Reza Zahiri Azar

Multi-channel pre-beamformed data acquisition system for research on advanced ultrasound imaging methods

The lack of open access to the pre-beamformed data of an ultrasound scanner has limited the research of novel imaging methods to a few privileged laboratories. To address this need, we have developed a pre-beamformed data acquisition (DAQ) system that can collect data over 128 array elements in parallel from the Ultrasonix series of research-purpose ultrasound scanners. Our DAQ system comprises three system-level blocks: 1) a connector board that interfaces with the array probe and the scanner through a probe connector port; 2) a main board that triggers DAQ and controls data transfer to a computer; and 3) four receiver boards that are each responsible for acquiring 32 channels of digitized raw data and storing them to the on-board memory. This system can acquire pre-beamformed data with 12-bit resolution when using a 40-MHz sampling rate. It houses a 16 GB RAM buffer that is sufficient to store 128 channels of pre-beamformed data for 8000 to 25000 transmit firings, depending on imaging depth; corresponding to nearly a 2-s period in typical imaging setups. Following the acquisition, the data can be transferred through a USB 2.0 link to a computer for offline processing and analysis. To evaluate the feasibility of using the DAQ system for advanced imaging research, two proof-of-concept investigations have been conducted on beamforming and plane-wave B-flow imaging. Results show that adaptive beamforming algorithms such as the minimum variance approach can generate sharper images of a wire cross-section whose diameter is equal to the imaging wavelength (150 μm in our example). Also, plane wave B-flow imaging can provide more consistent visualization of blood speckle movement given the higher temporal resolution of this imaging approach (2500 fps in our example).

Sub-sample displacement estimation from digitized ultrasound RF signals using multi-dimensional polynomial fitting of the cross-correlation function

A widely used time-domain technique for motion or delay estimation between digitized ultrasound RF signals involves the maximization of a discrete pattern-matching function, usually the cross-correlation. To achieve sub-sample accuracy, the discrete pattern-matching function is interpolated using the values at the discrete maximizer and adjacent samples. In prior work, only 1-D fit, applied separately along the axial, lateral, and elevational axes, has been used to estimate the sub-sample motion in 1-D, 2-D, and 3-D. In this paper, we explore the use of 2-D and 3-D polynomial fitting for this purpose. We quantify the estimation error in noise-free simulations using Field II and experiments with a commercial ultrasound machine. In simulated 2-D translational motions, function fitting with quartic spline polynomials leads to maximum bias of 0.2% of the sample spacing in the axial direction and 0.4% of the sample spacing in the lateral direction, corresponding to 38 nm and 1.31 μm, respectively. The maximum standard deviations were approximately 1% of the sample spacing in both the axial and the lateral directions, corresponding to 193 nm axially and 4.43 μm laterally. In simulated 1% axial strain, the same function fitting leads to mean absolute displacement estimation errors of 255 nm in the axial direction and 4.77 ?m in the lateral direction. In experiments with a linear array transducer, 2-D quartic spline fitting leads to maximum bias of 458 nm and 6.27 μm in the axial and the lateral directions, respectively. These results are more than one order of magnitude smaller than those obtained with separate 1-D fit when applied to the same data set. Simulations and experiments in 3-D yield similar results when comparing 3-D polynomial fitting with 1-D fitting along the axial, lateral, and elevational directions.

2-D high-frame-rate dynamic elastography using delay compensated and angularly compounded motion vectors: Preliminary results

This paper describes a new ultrasound-based system for high-frame-rate measurement of periodic motion in 2-D for tissue elasticity imaging. Similarly to conventional 2-D flow vector imaging, the system acquires the RF signals from the region of interest at multiple steering angles. A custom sector subdivision technique is used to increase the temporal resolution while keeping the total acquisition time within the range suitable for real-time applications. Within each sector, 1-D motion is estimated along the beam direction. The intraand inter-sector delays are compensated using our recently introduced delay compensation algorithm. In-plane 2-D motion vectors are then reconstructed from these delay-compensated 1-D motions. We show that Youngs modulus images can be reconstructed from these 2-D motion vectors using local inversion algorithms. The performance of the system is validated quantitatively using a commercial flow phantom and a commercial elasticity phantom. At the frame rate of 1667 Hz, the estimated flow velocities with the system are in agreement with the velocity measured with a pulsed-wave Doppler imaging mode of a commercial ultrasound machine with manual angle correction. At the frame rate of 1250 Hz, phantom Youngs moduli of 29, 6, and 54 kPa for the background, the soft inclusion, and the hard inclusion, are estimated to be 30, 11, and 53 kPa, respectively.

4-D x 3-D ultrasound: real-time scan conversion, filtering, and display of displacement vectors with a motorized curvilinear transducer

Recent research in the field of elastography has sought to expand displacement tracking to three dimensions. Once the 3-D volumes of displacement data have been obtained, they must be scan converted so that further processing, such as inversion methods to obtain tissue elasticity, can take place in Cartesian coordinates. This paper details an efficient and geometrically accurate algorithm to scan convert 3-D volumes of displacement vectors obtained from a motorized sector transducer. The proposed algorithm utilizes the physical scan geometry to convert the 3-D volumes of displacement data to both Cartesian coordinates and Cartesian displacements. Spatially varying filters are also proposed to prevent aliasing while minimizing data loss. Validation of the system has shown the algorithm to be correct to floating point precision, and the 3-D scan conversion and filtering can be performed faster than the native rate of data acquisition for the motorized transducer.

Comparison between 2-D cross correlation with 2-D sub-sampling and 2-D tracking using beam steering

We have previously presented multi-dimensional sub-sample motion estimation techniques that use multi-dimensional polynomial fitting to the discrete cross-correlation function to jointly estimate the sub-sample motion in all three spatial directions. Previous simulation and experimental results showed that these estimators significantly improve the performance of the motion estimation in 2-D and 3-D. In this short communication, we present additional simulation results and compare these techniques to 2-D tracking using beam steering. The results show that beam steering technique performs better in estimating the motion vector especially the lateral component.

Analysis of 2-D motion tracking in ultrasound with dual transducers.

We study displacement and strain measurement error of dual transducers (two linear arrays, aligned orthogonally and coplanar). Displacements along the beam of each transducer are used to obtain measurements in two-dimensions. Simulations (5MHz) and experiments (10MHz) are compared to measurements with a single linear array, with and without angular compounding. Translation simulations demonstrate factors of 1.07 larger and 8.0 smaller biases in the axial and lateral directions respectively, for dual transducers compared to angular compounding. As the angle between dual transducers decreases from 90° to 40°, for 1% compression simulations, the lateral RMS error ranges from 2.1 to 3.9μm compared to 9μm with angular compounding. Simulation of dual transducer misalignment of 1mm and 2° result in errors of less than 9μm. Experiments demonstrate factors of 3.0 and 5.2 lower biases for dual transducers in the axial and lateral directions respectively compared to angular compounding.

Importance of transducer position tracking for automated breast ultrasound: Initial assessments

Breast cancer causes the most cancer deaths in women all over the world. Given the unknown cause and invasive nature of the disease, early detection and treatment is crucial to the survival of the cancer patients. X-rays mammography is an effective screening tool for breast cancer but misses nearly half of them in women with dense breasts, in which case supplemental ultrasound (US) screening must be used. Recent development in 3D automated breast ultrasound (ABUS) provides a faster and potentially more accurate alternative to the conventional 2D hand-held US in detecting the tumors. Typically in these systems, a custom-built linear or curvilinear transducer is automatically swept over or rotated around the tissue to acquire a set of 2D images, from which a 3D high-resolution volume of the breast can be reconstructed. In this work, we study the importance of transducer position tracking in the ABUS volume reconstruction process. Our experiments show that an accurate localization of the images results in a more accurate volume reconstruction.

An automated breast ultrasound system for elastography

An automated breast ultrasound system (ABUS) has recently been introduced for breast screening and monitoring of cancer treatment. ABUS enables clinicians to acquire ultrasound images from the entire breast tissue in a few minutes. In this work we report the addition of tissue elasticity imaging to the ABUS system that enables the clinician to automatically acquire 3D strain volume of the breast tissue. The performance of the system is validated experimentally using a commercial breast elasticity phantom. The results show that the proposed automated system can reliably generate elasticity images of the breast tissue that can be reviewed by clinicians along side ultrasound images.

Real-time 1-D/2-D transient elastography on a standard ultrasound scanner using mechanically induced vibration

Transient elastography has been well established in the literature as a means of assessing the elasticity of soft tissue. In this technique, tissue elasticity is estimated from the study of the propagation of the transient shear waves induced by an external or internal source of vibration. Previous studies have focused mainly on custom single-element transducers and ultrafast scanners which are not available in a typical clinical setup. In this work, we report the design and implementation of a transient elastography system on a standard ultrasound scanner that enables quantitative assessment of tissue elasticity in real-time. Two new custom imaging modes are introduced that enable the system to image the axial component of the transient shear wave, in response to an externally induced vibration, in both 1-D and 2-D. Elasticity reconstruction algorithms that estimate the tissue elasticity from these transient waves are also presented. Simulation results are provided to show the advantages and limitations of the proposed system. The performance of the system is also validated experimentally using a commercial elasticity phantom.

Estimation of shear modulus ratio between inclusion and background using strain ratios in 2-D ultrasound elastography

The goal of this study is to assess the effects of region of interest (ROI) selection and lesion size on estimates of shear modulus ratio from strain ratios to quantify relative stiffness of breast tumors. A theoretical model and finite element method (FEM) simulations of lesions with various shear modulus ratios are created for a 2-D plane strain deformation. Both the lesion and the surrounding tissue are assumed to be linearly elastic, isotropic, homogenous, and incompressible. The results from the model and simulations are in agreement that the lesion-to-surrounding shear modulus ratio is linearly proportional to the axial normal strain ratio for small lesions when the ROI in the surrounding tissue is at least four lesion diameters away from the lesion. For larger lesions, FEM simulations show that the estimated strain ratio using the same ROI location increases with the lesion size and would overestimate the shear modulus ratio. Therefore, a correction factor is necessary for large breast lesions when strain ratios are used to estimate the shear modulus ratio. We also demonstrate that strain elastograms calculated using a speckle tracking method on simulated RF data are accurate enough to observe the same effect on strain ratio estimation. This result is confirmed using experimental data acquired from two tissue-mimicking phantoms. Our findings will help clinicians to estimate strain ratios and shear modulus ratios more accurately for more reliable comparison of one clinical examination to another.