Synthetic transmit aperture ultrasound (STAU) imaging can create images with as low as 2 emissions, making it attractive for 3D real-time imaging. Two are the major problems to be solved: (1) complexity of the hardware involved, and (2) poor image quality due to low signal to noise ratio (SNR). We have solved the first problem by building a scanner capable of acquiring data using STAU in real-time. The SNR is increased by using encoded signals, which make it possible to send more energy in the body, while reserving the spatial and contrast resolution. The performance of temporal, spatial and spatio-temporal encoding was investigated. Experiments on wire phantom in water were carried out to quantify the gain from the different encodings. The gain in SNR using an FM modulated pulse is 12 dB. The penetration depth of the images was studied using tissue mimicking phantom with frequency dependent attenuation of 0.5 dB/(cm MHz). The combination of spatial and temporal encoding have highest penetration depth. Images to a depth of 110 mm, can successfully be made with contrast resolution comparable to that of a linear array image. The in-vivo scans show that the motion artifacts do not significantly influence the performance of the STAU.

Conventional techniques used to design transducer apertures for medical ultrasound are generally iterative and ad-hoc. They do not guarantee optimization of parameters such as mainlobe width and sidelobe levels. We propose a dynamic aperture weighting technique, called the Minimum Sum Squared Error (MSSE) technique, that can be applied in arbitrary system geometries to design apertures optimizing these parameters. The MSSE technique utilizes a linear algebra formulation of the Sum Squared Error (SSE) between the point spread function (psf) of the system, and a goal or desired psf. We have developed a closed form expression for the aperture weightings that minimize this error and optimize the psf at any range. We present analysis for Continuous Wave (CW) and broadband systems, and present simulations that illustrate the flexibility of the technique.

A new method to increase the signal-to-noise-ratio (SNR) of synthetic transmit aperture (STA) imaging is investigated. The new approach is called temporally Encoded Multi-Element STA imaging (EMESTA). It utilizes multiple elements to emulate a single transmit element, and the conventional short excitation pulses are replaced by linear FM signals. Simulations using Field II and measurements are compared to linear array imaging. A theoretical analysis shows a possible improvement in SNR of 17 dB. Simulations are done using an 8.5 MHz linear array transducer with 128 elements. Spatial resolution results show better performance for EMESTA imaging after the linear array focus. Both methods have similar contrast performance. Measurements are performed using our experimental multi-channel ultrasound scanning system, RASMUS. The designed linear FM signal obtains temporal sidelobes below -55 dB, and SNR investigations show improvements of 4-12 dB. The depth performance is investigated using a multi-target phantom. Results show a 30 mm increase in penetration depth with improved spatial resolution. In conclusion, EMESTA imaging significantly increases the SNR of STA imaging, exceeding that of linear array imaging.

Traditionally, the number of transmit and receive processing channels is equal to the number of transducers (N) in an ultrasound imaging system. Certain applications limit the number of processing channels such that there are fewer channels than transducer elements. For these cases, a subset of M adjacent transducers-a multi-element subarray-performs echo transmission and reception. The processing channels are multiplexed across the array as beams are acquired from each of K subarrays. Combination of all subarray apertures creates a multi-element synthetic aperture (MSA) that represents the response of the entire system. Appropriate 1D filtering is applied in the spatial domain to restore a response approximating that of full phased array imaging. Compared to full phased array (FPA) imaging, MSA imaging reduces the number of front-end processing channels by a factor of N/M. Three variations of the method were simulated for a 128-element array using 32-element subarrays. The effects of the signal bandwidth, subsampling rate, and filter length on the reconstructed 2D point-spread functions are shown. The method closely approximates the performance of FPA imaging with fewer processing channels.

The pressure field emitted from multi-element medical ultrasound transducers can be simulated with Field II in the linear regime. By expanding this program's application to the nonlinear regime, beamforming schemes can be studied under strong focusing and high pressure levels as well, providing a valuable tool for simulating ultrasound harmonic imaging. An extended version of Field II is obtained by means of operator splitting. The pressure field is calculated by propagation of the field from the transducer through a number of planes. Every plane serves as a virtual aperture for the next plane, and nonlinear distortion is accounted for by the lossless Burgers' Equation. This method has no plane-wave approximation and the full effects of diffraction, attenuation, and nonlinear wave propagation can be observed under electronic focusing of array transducers in medical ultrasound systems. A single example of the approach is demonstrated by comparing results from simulations and measurements from a convex array transducer. The new simulation tool is capable of simulating the formation of higher harmonics in water on the acoustical axis. The generation of nonlinear higher harmonic components can be predicted with an accuracy of 2.6 dB and 2.0 dB for the second and third harmonic, respectively.

Frame rate in ultrasound imaging can be increased by simultaneous transmission of multiple beams using coded waveforms. However, the achievable degree of orthogonality among coded waveforms is limited in ultrasound, and the image quality degrades unacceptably due to interbeam interference. In this paper, an alternative combined time-space coding approach is undertaken. In the new method all transducer elements are excited with short pulses and the high time-bandwidth (TB) product waveforms are generated acoustically. Each element transmits a short pulse spherical wave with a constant transmit delay from element to element, long enough to assure no pulse overlapping for all depths in the image. Frequency shift keying is used for per element coding. The received signals from a point scatterer are staggered pulse trains which are beamformed for all beam directions and further processed with a bank of matched filters (one for each beam direction). Filtering compresses the pulse train to a single pulse at the scatterer position with a number of spike axial sidelobes. Cancellation of the ambiguity spikes is done by applying additional phase modulation from one emission to the next and summing every two successive images. Simulation results presented for QLFM and Costas spatial encoding schemes show that the proposed method can yield images with range sidelobes down to -45 dB using only two emissions.

Accurate measurement of tissue aberrations is necessary for effective adaptive ultrasound imaging. Higher order arrays provide more elements and a larger array footprint over which echo signals can be acquired. This allows for better sampling of the aberrator in both the azimuthal and elevation dimensions. These measured aberration profiles can then be used to correct the timing of transmitted and received RF signals to generate new images. We acquired single channel RF data on a 6.7 MHz, 8 x 128 array (Tetrad Co.) operating at F/1.0 in azimuth and F/2.9 in elevation. This array was interfaced to a Siemens Elegra scanner, allowing for data acquisition during routine phantom and clinical scanning. One-dimensional and two-dimensional physical near-field aberrators were used while imaging speckle only and spherical cyst-mimicking phantoms. In some experiments, neighboring elements were electronically tied in elevation to form ``taller'' elements. We collected individual channel data on each of 6 physical rows and then on a combination of rows to form 3x128, 2x128, and 1x128 arrays over a 6x128 aperture of the array. A least-mean-squares algorithm was employed to estimate the arrival time error induced by the tissue for the different array geometries. These aberration measurements were used to correct the images. In addition, point target simulations were performed to characterize the algorithm's performance for all four different array configurations. We present the performance of the adaptive imaging algorithm and discuss methods of combining arrival time profiles from axial and lateral tissue regions to improve adaptive imaging performance. Contrast results in simulation and phantom experiments with different aberrators are presented. We also discuss, in the context of our aberration measurement profiles, the array geometry requirements for successful adaptive imaging and the effects of the aberrators on sidelobe strength and contrast measurement. Results from performing adaptive imaging on clinical breast images using a 6x128 array geometry are also presented.

A 50 MHz annular array with six equal-area elements was designed. A novel fine-grain, 1-2 micron particle size, lead titanate was used as the piezoelectric material. The full set of electromechanical properties was calculated using resonance techniques. The low permittivity ((epsilon) ^s₃₃/(epsilon) ₀ equals 180) of the lead titanate allowed a larger overall diameter when matching to 50 ohm electronics, which kept the aspect ratio (width/height) of all the elements above 3. Laser micromachining was used to fully separate the array elements, and a double matching layer scheme was used to acoustically match the device. The array was modeled using both one-dimensional (KLM) and finite element modeling (PZFlex), and good agreement between the two was obtained. Using a broadband excitation, a center frequency of 49 MHz was obtained on the echo reflected from a flat reflector, with an inerstion loss of 20 dB and a bandwidth of 50%. Maximum calculated cross-talk values were below -30 dB between elements.

A promising alternative to piezoelectricity for high frequency array applications is optical generation and detection of ultrasound. An array element is defined by the size and location of a laser beam focused onto a suitable surface. We've built a two-dimensional synthetic receive array, where a HeNe laser probes the surface displacements of a thin reflective membrane. Using a conventional transducer as the ultrasound source, images with near optimal resolution and wide fields of view have been produced at 10 - 50 MHz. We are currently exploring a different form of optical detection where the incident ultrasound modulates the thickness of an etalon (a Fabry-Perot interferometer). Preliminary experiments demonstrate improved sensitivity using a high finesse etalon. Our work in optical generation of ultrasound uses the thermoelastic effect. A major drawback to thermoelastic generation has been poor conversion efficiency. We obtained an increase in conversion efficiency of nearly 20 dB using an optical absorbing film consisting of a mixture of polydimethylsiloxane (PDMS) and carbon black. Radiation pattern measurements indicate that we have produced a 75 MHz two-dimensional array element. These results demonstrate the potential of optoacoustic arrays for high frequency ultrasound imaging.

In breast cancer diagnosis, ultrasound examination provides useful additional diagnostic information. Moreover ultrasound does not harm biological tissue and can be applied frequently. But conventional ultrasound imaging methods lack both high spatial and temporal resolution. Usually, the scanner is operated manually and the tissue is deformed while getting as close as possible to the regions of interest. Therefore, image contents and image quality depend strongly on the operator. Exact measurement of tissue structures, like tumor size, is not possible. Instead of a manually controlled linear transducer array, we use ultrasound computer tomography (USCT) to image a volume directly. A few thousand ultrasound transducers are arranged in a cylindrical array around a tank containing the object to be examined coupled by water. Every single transducer is small enough to emit an almost spherical sound wave. While one transducer is transmitting, all others receive simultaneously. Afterwards a different transducer emits the next pulse. For volume reconstruction every transmitted, scattered and reflected signal is used. This new method allows reproducible image sequences with enhanced spatial and temporal resolution. For the benefit of more reconstructed 3D images per second, spatial resolution may be reduced offline. First tests with our prototype in a ring-shaped geometry have even showed nylon threads (0.4 mm) and an image quality superior to clinical ultrasound scanners.

This paper discusses the design, fabrication and testing of a 35 MHz linear ultrasonic array. The array features monolithic piezoelectric elements diced out of TRS 600FGHD fine grain high-density ceramic. A lossy urethane doped with gas filled microspheres is used as a kerf-filler to dampen inter-element acoustic propagation and reduce pulse length. The array incorporates a slotted single matching layer made from an unloaded epoxy. This matching layer also contributes to the reduction of pulse length and an increase in sensitivity. Array elements are spaced by a 50 mm pitch and interconnected via a flexible circuit. An 85 (Omega) transmission line coaxial cable is used to electrically match the array elements to the 50 (Omega) system electronics. The final 64-element array design is based on experimental results obtained from several four-element prototype arrays. An average center frequency of 34 MHz with a -6 dB bandwidth of at least 45% is achieved with the final prototype array. The maximum combined electrical and acoustical cross-talk for nearest and next nearest elements is less than -29 dB. The average -40 dB pulse length is 105 ns. The simple design and satisfactory performance of this array make it suitable for large-scale production.

Brachytherapy using small implanted radioactive seeds is becoming an increasingly popular method for treating prostate cancer. Seeds are inserted into the prostate transperineally using ultrasound guidance. Dosimetry software determines the optimal placement of seeds for achieving the prescribed dose based on ultrasonic determination of the gland boundaries. However, because of prostate movement after planning images are acquired and during the implantation procedure, seeds commonly are not placed in the desired locations and the delivered dose may differ from the prescribed dose. Current methods of ultrasonic imaging do not adequately display implanted seeds for the purpose of correcting the delivered dose. We are investigating new methods of ultrasonic imaging that overcome limitations of conventional ultrasound. These methods include resonance, modified elastographic, and signature techniques. Each method shows promise for enhancing the visibility of seeds in ultrasound images. Combining the information provided by each method may reduce ambiguities in determining where seeds are present or absent. If successful, these novel imaging methods will enable correction of seed-misplacement errors during the implantation procedure, and hence will improve the therapeutic radiation dose delivered to target tissues.

Myocardial structural changes caused by infarction / reperfusion may result in increased scatterer density and variation of scatterer arrangement in ultrasound imaging. Homodyned K (HD_K) distribution is employed in this paper to model the backscattered signal from both normal and reperfused infarcted myocardium and is used to characterize them. Statistical testing showed that among the Rayleigh, K, Nakagami and Homodyned K distributions, the Homodyned K distribution is the best model to describe ultrasound signal backscattered from both normal and infarcted reperfused myocardium. Using HD_K distribution, in vivo demodulated RF data (8.5MHz) from anterior myocardial wall, as imaged from both left and right ventricle at baseline and after infarction/reperfusion, were analyzed. Significant increase of scatterer density in reperfused infarcted myocardium has been found compared to the normal myocardium. We concluded that HD_K distribution has potential to distinguish reperfused infarcted myocardium from normal using high frequency ultrasound imaging, irrespective of LV or RV data acquisition.

Objective: To investigate and diagnose testicular pathology in patients with testicular dysfunction using the technique of ultrasound power spectrum analysis. Methods: Testicular ultrasound studies with power spectrum tissue characterization analysis were performed on men with testicular abnormalities as well as normal controls. Semen analysis, biopsy data, microscopic intra-operative findings and data pertaining to testicular function were collected for each surgically evaluated subject. Ultrasound data were analyzed for power spectrum characteristics of microscopic scatterer size and concentration within discrete areas of testicular tissue. Results: Patients with varicoceles and greater than 2x10⁶ sperm/ml on semen analysis had larger average scatterer size (107.7 micrometers ) and lower scatterer concentration (-15.02 dB) than non-obstructed, azoospermic patients with varicoceles (92.4 micrometers and -11.41 dB, respectively). Subjects with obstructed azoospermia had slightly larger average tissue scatterer size (108.1 micrometers ) and lower concentration (-15.73 dB) while normal control data revealed intermediate values of size (102.3 micrometers ) and concentration (-13.1 dB) of scatterers. Spectral data from pure testicular seminoma lesions had the lowest average scatterer size (82.3 micrometers ) with low relative concentration (-14.7 dB). Summary: Ultrasound tissue characterization based on RF spectrum analysis may distinguish different types of testicular pathology including obstructed and non-obstructed azoospermia and tissue changes due to varicocele and tumor.

Eight retired breeder rats were acquired that had developed spontaneous mammary tumors. Tumors were diagnosed microscopically as mammary gland fibroadenomas. Two-dimensional gray-scale B-mode images of the tumors in the rats were constructed from backscattered echoes using an 8 MHz (90% bandwidth) single element ultrasonic transducer. From the gray-scale B-mode images, regions-of-interest (ROIs) were selected in the tumors and surrounding tissues. The power spectra of backscattered RF echoes gated from the ROIs were used to estimate the average scatterer diameters and concentrations. A unique estimation scheme was used to obtain the average scatterer diameters and concentrations. The average scatterer diameter was related to the slope of the best-fit line to the reduced measured power spectrum versus the frequency squared. The scatterer concentration was determined from the intercept of the best-fit line. The reduced measured power spectrum is the measured power spectrum minus 40 log of the frequency. Parametric B-mode images were constructed by converting ROI boxes into colored pixels. The color of the pixels was related to the estimated scatterer properties. The images showed a distinct difference between the tumor and surrounding healthy tissues. Scatterer sizes inside the tumor were on average 30% larger than scatterer sizes in surrounding normal tissues.

Purpose: While histopathologic evaluation of ocular melanoma permits assessment of metastatic risk, this is not possible if visual function is to be preserved. In this report, we describe ultrasound methods for non-invasive evaluation of metastatic risk. Methods: Radiofrequency (RF) ultrasound data were acquired prior to enucleation in 117 eyes with untreated malignant melanoma. Extracellular matrix patterns, an indicator of metastatic potential, were identified in histologic sections. We determined calibrated backscatter power spectra, estimates of effective scatterer size and concentration, and the percentage of tumor area comprised of PAS-positive patterns in the anterior, posterior and core regions of the tumors. We compared the spatial correlation of histologic and acoustic properties, performed linear discriminant analysis to define prognostic models, and used receiver operating characteristic (ROC) curves to evaluate models. Results: Statistically significant correlations were found between acoustic parameters and PAS-positive patterns, although their spatial distributions were only weakly related. Stepwise linear discriminant analyses produced models with three to five variables, and ROC areas as high as 0.89. Conclusion: Acoustic spectrum analysis provides information not evident in conventional gray-scale ultrasonograms regarding tissue microstructure. Our results confirm a relationship between spectra and the presence of extracellular matrix patterns associated with metastatic risk.

We are developing ultrasonic strain-flow imaging instrumentation specifically to facilitate improved diagnosis of breast tissue disease. An 8-ring, 30-mm-diameter, f/1.5, spherically focused annular array was built to generate broadband, 10 MHz pulses at a rate up to 10 KHz. This transducer uses the synthetic receive aperture technique to record echoes while being mechanically steered by a linear positioner under microprocessor control. Specifically, a 200Vpp signal is applied to all rings simultaneously on transmission. Individual rings are then sequentially multiplexed to a receiver. Echoes are dynamically delayed and coherently summed off-line to adjust the receive focus and extend the depth of focus. The aperture material is a 1-3 composite built by Imasonic SA, Besancon, France. One advantage of the design is that it provides a well-focused axisymmetric beam with an improved depth of focus to acquire images at a high spatial resolution. Pulse-echo simulations of our array using the Field II software package show a -6dB beam width at 0.24 mm and -6dB depth of focus at 2.4 mm that can be extended to 3.2 mm with dynamic focusing. These simulations agree with later measurements performed on the transducer. Our flexible aperture design allows us to drop outer rings significantly increasing the depth of focus (up to 56% increase by dropping 3 rings) with a tolerable decrease in lateral resolution (27% increase in beam width). We expect that our probe will enable us to examine detailed biological processes throughout the malignant growth period of a tumor tissue that exhibits elastic anisotropy, thus providing high resolution ultrasound images over an extended and adjustable depth of field.

We describe a novel strain estimation method, which is capable of accurate strain estimation in the presence of large or irregular tissue motion. In conventional elastography, tissue strains induced by external compression applied to the tissue surface, are estimated by cross-correlation analysis of echo signals obtained before and after compression. Large and irregular tissue motions significantly change echo-signal shapes, causing major echo-signal decorrelations. In the presence of significant signal decorrelations, the estimated displacements may have many discontinuities because of false peak errors (primary correlation peak smaller than a secondary peak). However, because tissue is virtually incompressible, true elastographic displacements have spatial continuity in both axial and lateral directions, and the true correlation peaks, albeit diminished, are still present and detectable in the presence of spurious peaks. Our approach treats the total ensemble of correlation functions vs. depth and use information from neighboring areas to remove ambiguity that results from false peak errors. Preliminary results using finite-element simulation show that the correlation-tracking strain estimator (CTSE) can produce excellent strain estimates in harsh environments.

A technique to improve detection of brachytherapy seeds in ultrasound images is presented. Seeds modified to include a small ferrous or magnetic component are vibrated with an amplitude of a few microns within the tissue by an external magnetic field. The vibration is detected by standard Power Doppler or Pulse Wave Doppler systems, which pinpoint the source of vibration and thus the seed within the image. The results of in vitro experiments on agar and liver-tissue phantoms are presented to demonstrate the feasibility of the method. MRI compatibility issues are addressed.

Conventional ultrasound Doppler velocity measurements are scaled by the cosine of the angle between the blood flow axis and ultrasound beam axis. In the approach used here, a transducer array was used to acquire a first cross-sectional Doppler data set of the vessel under examination. The transducer array was then moved to a different angle to acquire a second cross-sectional Doppler data set. Thereafter, we used the known angle between the two arrays ultrasound beams and the cosine (theta) scaled Doppler estimates to solve for the true angle between the blood flow axis and ultrasound beam axis of the first data set. Upon integrating the angle corrected velocity estimates over the entire vessel cross-section, we were able to estimate blood volume flow rate. The performance of the new approach was tested in a flow phantom that was designed to provide a constant flow in a simulated vessel. The data were collected for two sets of angles and three different flow velocities for each angle set. The unknown Doppler angle was calculated from the data and used to correct the flow velocity.

A new ultrasound technique for determining three-dimensional velocity vectors has been devised using radio frequency (RF) data from commercially available scanners. Applied to blood flow, this technique could prove useful for evaluating hemodynamics and detecting stenoses. Three orthogonal velocity vectors are computed from the RF signals of two steered beams from a single array. The in-plane velocities are determined using standard Doppler analysis, while the out-of-plane component is derived from the total velocity as computed from temporal decorrelation and the in-plane components. The technique was tested using contrast agent pumped through a flow tube. A GE Vingmed SystemV scanner with a 10 MHz linear array provided scans at beam steering angles of +/- 20 degree(s). Both Doppler velocities and temporal complex decorrelation were computed for each digitized voxel. Additional studies were done on a blood mimicking fluid and in vivo with a canine femoral artery. Vector plots were constructed to show flow for various transducer angles. Angle estimates were within 20 degree(s), and the mean error for the velocity amplitude was less than 15%. The in vivo results provided velocity estimates consistent with the literature. The proposed method, unlike current Doppler velocity measurement techniques, provides quantitative velocity information independent of transducer orientation.

Several strategies, known as clutter or wall Doppler filtering, were proposed to remove the strong echoes produced by stationary or slow moving tissue structures from the Doppler blood flow signal. In this study, the matching pursuit (MP) method is proposed to remove clutter components. The MP method decomposes the Doppler signal into wavelet atoms that are selected in a decreasing energy order. Thus, the high-energy clutter components are extracted first. In the present study, the pulsatile Doppler signal s(n) was simulated by a sum of random-phase sinusoids. Two types of high-amplitude clutter signals were then superimposed on s(n): a time-varying low frequency component (type 1), covering systole and early diastole, and short transient clutter signals (type 2), distributed within the whole cardiac cycle. The Doppler signals were modeled with the MP method and the most dominant atoms were subtracted until the signal-to-clutter (S/C) ratio reached a maximum. For the type 1 clutter signal, the improvement in the S/C ratio was 19.0 +/- 0.6 dB, and 72.0 +/- 4.5 atoms were required to reach this performance. For the transient type 2 clutter signal, exactly 10 atoms were required and the maximum improvement in S/C ratio was 5.5 +/- 0.5 dB. These results suggest the possibility of using this signal processing approach to implement clutter rejection filters on ultrasound commercial instruments.

Observing the correlation between the information in neighboring resolution cells, we propose a methodology to allow better analysis of Doppler ultrasound signals. A blind source separation problem is formulated to discern the different signal components for correct interpretation of the data using independent component analysis (ICA). The Doppler signal is modeled as the summation of the true velocity signal, baseline fluctuation, and random noise. The baseline fluctuation component can be considered as a deterministic yet unknown signal. A simple adaptive denoising technique is applied to reduce the effective dimension of the noise subspace. Then, given a region of interest, the temporal signals corresponding to all pixels within this region undergo the ICA iteration to compute a set of independent signals that most represent the actual components present within the data. Subsequently, a comparison of the temporal variations of these signals allows the user identify the components of the signals that correspond to baseline variation or random noise manually or using semi-automated techniques. The new technique shows large potential to alleviate some of the limitations in this demanding imaging mode as well as to make the interpretation of the results more robust.

Inappropriate blood coagulation plays an important role in diseases including stroke, heart attack, and deep vein thrombosis (DVT). DVT arises when a blood clot forms in a large vein of the leg. DVT is detrimental because the blood flow may be partially or completely obstructed. More importantly, a potentially fatal situation may arise if part of the clot travels to the arteries in the lungs, forming a pulmonary embolism (PE). Characterization of the mechanical properties of DVT could improve diagnosis and suggest appropriate treatment. We are developing a technique to assess mechanical properties of forming thrombi. The technique uses acoustic radiation force as a means to produce small, localized displacements within the sample. Returned ultrasound echoes are processed to estimate the time dependent displacement of the sample. Appropriate mechanical modeling and signal processing produce plots depicting relative mechanical properties (relative elasticity and relative viscosity) and force-free parameters (time constant, damping ratio, and natural frequency). We present time displacement curves of blood samples obtained during coagulation, and show associated relative and force-free parameter plots. These results show that the Voigt model with added mass accurately characterizes blood behavior during clot formation.

This research aims at developing a three-dimensional (3D) ultrasound system for carotid and brachial artery scanning for evaluating vessel wall characteristics. In the long term, we seek to test hypothesis that the artery wall measurements of carotid intima-media-thickness and brachial flow mediated dilatation using 3D ultrasound data provide better repeatability than those derived from conventional 2D ultrasound scans. The approach is to implement a free-hand data acquisition scheme using conventional 2D medical ultrasound scanner, develop data processing algorithms for appropriately registering and displaying the volumetric ultrasound vessel scans, and develop techniques for measuring vessel wall characteristics. The system uses electromagnetic sensor mounted on the transducer to acquire position and orientation of each image slice as the transducer is moved freely to scan the area of interest. These non-parallel images are registered into a 3D dataset for reconstruction, segmentation, and measurements of the vessel wall structure. A simple calibration object is developed using a small stainless-steel sphere in a fixed position to perform coordinate transformations and pixel registration. A commercial 3D ultrasound tissue-mimicking phantom is used for assessment of freehand 3D data acquisition, calibration, registration, and visualization schemes. Early results of experimental carotid artery scans of volunteers are presented.

This paper provides a method for imaging a target scene in space and velocity domains. The method relies on a transmit focusing method to spotlight the region of interest with an ultrasonic beam; this is to improve the energy of the echo signature from the desired area. A receiver array is used to measure the resultant echoed signals; the receiver array provides spatial resolution within the spotlights target area. A wide-bandwidth continuous wave (WB-CW) signaling scheme is used to resolve targets in the Doppler domain.

Three-dimensional (3D) intravascular ultrasound provides valuable insight into the tissue characteristics of the coronary wall and plaque composition. However, artifacts due to cardiac motion and vessel wall pulsation limit the accuracy and variability of coronary lumen and plaque volume measurement in 3D IVUS images. ECG-gated image acquisition can overcome these artifacts but results in lengthy acquisition times. Our goal is to reconstruct a 3D IVUS image with negligible vessel pulsation artifacts, by developing an image-based retrospective gating method to track 2D IVUS images belonging to the same cardiac phase. Our approach involves selecting 2D IVUS images belonging to the same cardiac phase from an asynchronously acquired series, by tracking the changing lumen contour over the cardiac cycle. The algorithm was tested using a custom-built coronary phantom and on patient images. 3D non-gated and gated IVUS images were assembled and compared. The extent of pulsation artifacts in the 3D images was estimated by measuring the standard deviation in the shift in the position of the lumen boundary in each cross-sectional slice over the 3D IVUS image. A reduction in pulsation artifact of over 97% was observed in the 3D image assembled using our method.

In medical ultrasonography, speckle model parameters are dependent on scatterer density and regularity, and can be exploited for use in tissue characterization. The purpose of the current study is to quantify the goodness-of-fit of two models (the Nakagami and K distributions), applied to envelope data representing a range of clinically relevant scattering conditions. Ground truth data for computing goodness-of-fit were generated with envelope simulators. In the first simulation, 100 datasets of various sample sizes were generated with 40 scatterer densities, ranging from 0.025 to 20. Kolmogorov-Smirnov significance values quantified the goodness-of-fit of the two models. In the second simulation, densities ranged from 2 to 60, and additional scattering parameters were allowed to vary. Goodness-of-fit was assessed with four statistical tests. Although the K distribution has a firm physical foundation as a scattering model, inaccuracy and high standard deviation of parameter estimates reduced its effectiveness, especially for smaller sample sizes. In most cases, the Nakagami model, whose parameters are relatively easy to compute, fit the data best, even for large scatterer densities.

We have developed a family of objective features in order to provide non-invasive, reliable means of distinguishing benign from malignant breast lesions. These include acoustic features (echogenicity, heterogeneity, shadowing) and morphometric features (area, aspect ratio, border irregularity, margin definition). These quantitative descriptors are designed to be independent of instrument properties and physician expertise. Our analysis included manual tracing of lesion boundaries and adjacent areas on grayscale images generated from RF data. To derive quantitative acoustic features, we computed spectral parameter maps of radio-frequency (RF) echo signals (calibrated with system transfer function and corrected for diffraction) within these areas. We quantified morphometric features by geometric and fractal analysis of traced lesion boundaries. Although no single parameter can reliably discriminate cancerous from non-cancerous breast lesions, multifeature analysis provides excellent discrimination of cancerous and non-cancerous lesions. RF echo-signal data used in this study were acquired during routine ultrasonic examinations of biopsy-scheduled patients at three clinical sites. Our data analysis for 130 patients produced an ROC-curve area of 0.9164 +/- 0.0346. Among the quantitative descriptors, lesion heterogeneity, aspect ratio, and a border irregularity descriptor were the most useful; some morphometric features (such as the border irregularity descriptor) were particularly effective in lesion classification.

Support Vector Machines are a general algorithm based on guaranteed risk bounds of statistical learning theory. They have found numerous applications, such as in classification of brain PET images, optical character recognition, object detection, face verification, text categorization and so on. In this paper we propose the use of support vector machines to segment lesions in ultrasound images and we assess thoroughly their lesion detection ability. We demonstrate that trained support vector machines with a Radial Basis Function kernel segment satisfactorily (unseen) ultrasound B-mode images as well as clinical ultrasonic images.

New ultrasound data, obtained with a circular experimental scanner, are compared with data obtained with standard X-ray CT. Ultrasound data obtained by scanning fixed breast tissue were used to generate images of sound speed and reflectivity. The ultrasound images exhibit approximately 1 mm resolution and about 20 dB of dynamic range. All data were obtained in a circular geometry. X-ray CT scans were used to generate X-ray images corresponding to the same 'slices' obtained with the ultrasound scanner. The good match of sensitivity, resolution and angular coverage between the ultrasound and X-ray data makes possible a direct comparison of the three types of images. We present the results of such a comparison for an excised breast fixed in formalin. The results are presented visually using various types of data fusion. A general correspondence between the sound speed, reflectivity and X-ray morphologies is found. The degree to which data fusion can help characterize tissue is assessed by examining the quantitative correlations between the ultrasound and X-ray images.

In this paper we introduce a new speckle suppression technique for medical ultrasound images that incorporates morphological properties of speckle as well as tissue classifying parameters. Each individual speckles is located, and, exploiting our prior knowledge on the tissue classification, it is determined whether this speckle is noise or a medically relevant detail. We apply the technique on images of neonatal brains affected by White Matter Damage (leukomalacia). The results show that applying an active contour on a processed image, in order to segment the affected areas, yields a segmentation much closer to that of an expert.

Ultrasound speckle carries information about the interrogated scattering microstructure. The complex signal is represented as a superposition of signals due to all scatterers within a resolution cell volume, V_E. A crossbeam geometry with separate transmit and receive transducers is well suited for such studies. The crossbeam volume, V_E is defined in terms of the overlapping diffraction beam patterns. Given the focused piston transducer's radius and focal distance, a Lommel diffraction formulation suitable for monochromatic excitation is used to calculate V_E as a function of frequency and angle. This formulation amounts to a Fresnel approximation to the diffraction problem and is not limited to the focal zone or the far field. Such diffraction corrections as V_E are needed to remove the system effects when trying to characterize material using moment analysis. Theoretically, V_E is numerically integrated within the overlapping region of the product of the transmit-receive transfer functions. Experimentally, V_E was calculated from the field pattern of a medium-focused transducer excited by a monochromatic signal detected by a 0.5mm diameter PVDF membrane hydrophone. We present theoretical and experimental evaluations of V_E for the crossbeam geometry at frequencies within the transducers' bandwidth, and its application to tissue microstructure characterization.

Ultrasound images, as a special case of coherent images, are normally corrupted with multiplicative noise i.e. speckle noise. Speckle noise reduction is a difficult task due to its multiplicative nature, but good statistical models of speckle formation are useful to design adaptive speckle reduction filters. In this article a new statistical model, emerging from the Multiplicative Model framework, is presented and compared to previous models (Rayleigh, Rice and K laws). It is shown that the proposed model gives the best performance when modeling the statistics of ultrasound images. Finally, the parameters of the model can be used to quantify the extent of speckle formation; this quantification is applied to adaptive speckle reduction filter design. The effectiveness of the filter is demonstrated on typical in-vivo log-compressed B-scan images obtained by a clinical ultrasound system.

In contrast to standard reflection ultrasound (US), transmission US holds the promise of more thorough tissue characterization by generating quantitative acoustic parameters. We compare results from a conventional US scanner with data acquired using an experimental circular scanner operating at frequencies of 0.3 - 1.5 MHz. Data were obtained on phantoms and a normal, formalin-fixed, excised breast. Both reflection and transmission-based algorithms were used to generate images of reflectivity, sound speed and attenuation.. Images of the phantoms demonstrate the ability to detect sub-mm features and quantify acoustic properties such as sound speed and attenuation. The human breast specimen showed full field evaluation, improved penetration and tissue definition. Comparison with conventional US indicates the potential for better margin definition and acoustic characterization of masses, particularly in the complex scattering environments of human breast tissue. The use of morphology, in the context of reflectivity, sound speed and attenuation, for characterizing tissue, is discussed.

Extremely high quality data was acquired using an experimental ultrasound scanner developed at Lawrence Livermore National Laboratory using a 2D ring geometry with up to 720 transmitter/receiver transducer positions. This unique geometry allows reflection and transmission modes and transmission imaging and quantification of a 3D volume using 2D slice data. Standard image reconstruction methods were applied to the data including straight-ray filtered back projection, reflection tomography, and diffraction tomography. Newer approaches were also tested such as full wave, full wave adjoint method, bent-ray filtered backprojection, and full-aperture tomography. A variety of data sets were collected including a formalin-fixed human breast tissue sample, a commercial ultrasound complex breast phantom, and cylindrical objects with and without inclusions. The resulting reconstruction quality of the images ranges from poor to excellent. The method and results of this study are described including like-data reconstructions produced by different algorithms with side-by-side image comparisons. Comparisons to medical B-scan and x-ray CT scan images are also shown. Reconstruction methods with respect to image quality using resolution, noise, and quantitative accuracy, and computational efficiency metrics will also be discussed.

The Doppler velocimeter developed allows to determine the angle between the ultrasonic beam and the velocity vector of the flow, and to calculate the precise blood flow in a vessel. Four piezoelectric transducers constitute the Doppler velocimeter. Three of these transducers are positioned to form an equilateral triangle (base of a pyramid). When these transducers move simultaneously, backward or forward from the initial position, the emitted ultrasonic beams focalize on a position (peak of the pyramid) closer or farther from the transducers faces, according to the depth of the vessel where we intend to measure de flow. The angle between the transducers allows adjusting the height of this pyramid and the position of the focus (where the three beams meet). A forth transducer is used to determine the diameter of the vessel and monitor the position of the Doppler velocimeter relative to the vessel. Simulation results showed that with this technique is possible to accomplish precise measurement of blood flow.

As a tool for cardiac assessment, 3D echocardiography (3DE) has largely been limited to use by experts capable of qualitative determination of left ventricular (LV) function. The usefulness of 3DE can be extended to physicians in critical care settings who have minimal training in echocardiography if it delivers quantitative parameters of LV function without expert supervision. As a critical step to generate the quantitative measures, we develop an algorithm that automatically locates and tracks the LV boundary through a sequence of 2D frames, such as those constituting a 3DE data set. A novel approach of the algorithm in this paper is the computation of an edge probability field for each frame. For unsupervised processing, the algorithm incorporates a template-based search in the first frame using prior knowledge about the LV shape. Then, by active contour (or snake) matching, the LV boundary is obtained as a MAP estimate. The internal energy constraints of the snake serve as the prior probability. Unlike the conventional snakes that depend on hard edge information as external energy, the snake in our approach uses the soft information in the edge probability field. We also can obtain a measure of confidence in the boundary estimate from the value of the energy function at the estimated contour. The estimate from one frame is used to initialize active contour matching in the next frame for LV tracking.

In medical ultrasound signal-to-noise ratio improvements of approximately 15-20 dB can be achieved by using coded waveforms. Exciting the transducer with an encoded waveform necessitates compression of the response which is computationally demanding. This paper investigates the possibility of reducing the workload without introducing errors. N_e - 1 compression filtrations (convolutions) can be saved by inverting the precedence of compression and beamforming (called post-compression), when N is the number of transducer elements. Post-compression with dynamic receive focusing will theoretically introduce errors. Simulations and measurements show that increasing the depth of the scatterers results in a decreased error. Transmit focus depth and the distance between focus points have a significant influence on the error. The size of the error is studied and a new scheme for correcting the error is proposed. The study is done by simulations in Field II and by measurements with our experimental scanner RASMUS. The measurements are done on a string phantom and in-vivo on the abdomen of a male volunteer.

In this paper, we propose an efficient algorithm for organ recognition in ultrasound images using log power spectrum. The main procedure of the algorithm consists of feature extraction and feature classification. In the feature extraction, as a translation invariant feature, log power spectrum is used for extracting the information on the echo of organ tissues from a preprocessed input image. In the feature classification, Mahalanobis distance is used as a measure of the similarity between the feature of an input image and the representative feature of each class. Experimental results for real ultrasound images show that the proposed algorithm yields the maximum 30% improvement of recognition rate over the recognition algorithm using power spectrum and Euclidean distance, and results in 10-40% improvement of recognition rate over the recognition algorithm using weighted quefrency complex cepstrum.

This paper investigates the concept of virtual source elements. It suggests a common framework for increasing the resolution, and penetration depth of several imaging modalities by applying synthetic aperture focusing (SAF). SAF is used either as a post focusing procedure on the beamformed data, or directly on the raw signals from the transducer elements. Both approaches increase the resolution. The paper shows that in one imaging situation, there can co-exist different virtual sources for the same scan line - one in the azimuth plane, and another in the elevation. This property is used in a two stage beamforming procedure for 3D ultrasound imaging. The position of the virtual source, and the created waveform are investigated with simulation, and with pulse-echo measurements. There is good agreement between the estimated wavefront and the theoretically fitted one. Several examples of the use of virtual source elements are considered. Using SAF on data acquired for a conventional linear array imaging improves the penetration depth for the particular imaging situation from 80 to 110 mm. The independent use of virtual source elements in the elevation plane decreases the respective size of the point spread function at 100 mm below the transducer from 7mm to 2 mm.

Ultrasound phantoms are used as a quality assurance mechanism for assessing the imaging performance of ultrasound transducers. Several phantoms have been fabricated specifically for imaging by means of high frequency transducers with a center frequency ranging from 20 to 100 MHz. To quantify the transducers imaging performance the spatial resolution, dead zone, linear fidelity, depth of penetration and image uniformity are measured from ultrasound images created by scanning specially designed phantoms. Eight micron diameter tungsten (high acoustic reflectivity and diameter size less than (lambda) /2) wire targets are used for all the phantoms. Transducer characterization consists of a standard pulse echo analysis and insertion loss measurement for each transducer. Imaging of quality assurance phantoms and transducer characterization provide a practical means for evaluating the performance of high frequency transducers.

Tomographic images of tissue phantoms and a sample of breast tissue have been produced from an acoustic synthetic array system for frequencies near 500 kHz. The images for sound speed and attenuation show millimeter resolution and demonstrate the feasibility of obtaining high-resolution tomographic images with frequencies that can deeply penetrate tissue. The image reconstruction method is based on the Born approximation to acoustic scattering and is a simplified version of a method previously used by Andre (Andre, et. al., Int. J. Imaging Systems and Technology, Vol 8, No. 1, 1997) for a circular acoustic array system. The images have comparable resolution to conventional ultrasound images at much higher frequencies (3-5 MHz) but with lower speckle noise. This shows the potential of low frequency, deeply penetrating, ultrasound for high-resolution quantitative imaging.

Sonoelastography is a vibration Doppler technique for imaging the relative elasticity of tissues. Detectability of hard lesions of various sizes has previously been demonstrated in tissue phantoms by our group. Because real tissue differs from phantom material, the injection of formaldehyde in fresh liver tissue is being used as an in-vitro lesion model. Pieces of fresh calf liver were embedded in an agar gel then injected with a bolus of 37% formaldehyde to create a stiff lesion. Two and three-dimensional sonoelastography and b-mode images were acquired. The lesions were visible in each sonoelastography image as a region of reduced vibration. After imaging, lesions were dissected and measured for size and volume. One 0.4 cc bolus injection of formaldehyde created a lesion with a volume of 10.3 cc in the sonoelastography image compared to 9.3 cc using fluid displacement of the dissected lesion. A 0.33 cc injection of formaldehyde lesion created a volume of 5 cc in the sonoelastography image compared to 4.4 cc using fluid displacement. Sonoelastography imaging techniques for imaging hard lesions in phantoms can be successfully extended to imaging formaldehyde induced lesions in real tissue.

In this study, we developed a faster and more accurate method for the detection of corneal thickness using an ultrasonic pulse-echo method. After a ultrasound pulse was transmitted to the cornea, the echo was received in radio frequency. The corneal thickness was calculated from the time interval between the two echoes reflected at front and rear interfaces of the cornea. The time interval between two echoes was obtained by detecting the peaks of the digitized echoes. A phase-adjusting method, which is superior to the often-used interpolation approach, was used to improve the time resolution. Based on this method, a MCS-8031 microprocessor based pachymeter for measuring corneal thickness was developed. The center frequency of the ultrasonic transducer works was 30MHz and the echoes were digitized at a 180MHz sampling frequency. The device showed an accuracy of 1 micrometers for 10 repeated measurements.

3D spatial compounding involves the combination of two or more 3D ultrasound (US) data sets, acquired under different insonation angles and windows, to form a higher quality 3D US data set. An important requirement for this method to succeed is the accurate registration between the US images used to form the final compounded image. We have developed a new automatic method for rigid and deformable registration of 3D US data sets, acquired using a freehand 3D US system. Deformation is provided by using a 3D thin-plate spline (TPS). Our method is fundamentally different from the previous ones in that the acquired scattered US 2D slices are registered and compounded directly into the 3D US volume. Our approach has several benefits over the traditional registration and spatial compounding methods: (i) we only peform one 3D US reconstruction, for the first acquired data set, therefore we save the computation time required to reconstruct subsequent acquired scans, (ii) for our registration we use (except for the first scan) the acquired high-resolution 2D US images rather than the 3D US reconstruction data which are of lower quality due to the interpolation and potential subsampling associated with 3D reconstruction, and (iii) the scans performed after the first one are not required to follow the typical 3D US scanning protocol, where a large number of dense slices have to be acquired; slices can be acquired in any fashion in areas where compounding is desired. We show that by taking advantage of the similar information contained in adjacent acquired 2D US slices, we can reduce the computation time of linear and nonlinear registrations by a factor of more than 7:1, without compromising registration accuracy. Furthermore, we implemented an adaptive approximation to the 3D TPS with local bilinear transformations allowing additional reduction of the nonlinear registration computation time by a factor of approximately 3.5. Our results are based on a commercially available tissue-mimicking abdominal phantom and in-vivo muscle data.

The Born-approximation deconvolved inverse scattering (BADIS) imaging technique is an alternative to the conventional pulse-echo method, which presents the magnitude of the reflected pulse as the distribution of the target acoustic reflectivity. BADIS deconvolves the incident pulse from the reflected pulse, and uses the resulting impulse response to produce an image of the acoustic impedance distribution. It is applicable mainly to structures that resemble a layered medium. The images captured by this method prove to have improved resolution and are free of speckle. With the BADIS method one can use ultrasound of lower frequencies than would be required by the pulse-echo method to achieve the same resolution. To provide further improvement of images the second harmonic signals can be employed. Here we describe the combined BADIS-Harmonic Imaging method. For this purpose the hybrid transducer by Krautkramer Branson Co., which consists of a cylindrical 5 MHz transducer wrapped in an annulus-shaped 2.25 MHz transducer, has been used. The image phantom was a two- plastic film structure with drilled holes in the top layer. It is shown that the BADIS analysis of the second harmonic reflection data provides improved images.