This talk will first give a general discussion on the ultrasound media characteristics of blood and spectral densities of tissues. The first-order scattering theory, multiple scattering theory, Doppler spectrum, cw and pulse scattering, focused beam, beam spot-size, speckle, texture, and rough interface effects will be presented. Imaging through tissues will then be discussed in terms of temporal and spatial resolutions, contrast, MTF (modulation transfer function), SAR and confocal imaging techniques, tomographic and holographic imaging, and inverse scattering. Next, we discuss optical diffusion in blood and tissues, radiative transfer theory, photon density waves, and polarization effects.

Ultrasound contrast agents are bubbles, 1-10 microns in radius, encapsulated by a lipid, protein, polymer or fluid shell. The agents have been used to distinguish the acoustic scattering signatures of blood from those of the surrounding tissue. This is possible due to the nonlinear response of the agent, which is similar to that of a free gas bubble. Upon sufficient forcing the agents will oscillate nonlinearly about their equilibrium radius, and for specific conditions, produce nonlinear resonance responses which are integer multiples of the primary resonance. Ultrasound tissue perfusion studies have been developed which are based on the destruction of contract agents coupled to the measurement of blood flow. Nevertheless, many outstanding issues remain in contrast agent design especially with respect to emerging applications. Even with the use of higher order harmonics there is a lack of an acoustic signature or destruction mechanism at frequencies above approximately 5.0 MHz with conventional agents. The design and use of a high frequency contrast agent is addressed by exploiting the multiple scattering response of agents modled as spherical elastic shells. Also considered is the nonlinear response of elastic-shelled agents. The considerations of shells modeled as linear and nonlinear elastic materials are discussed. The use of contrast agents for targeted drug delivery has recently received much attention. More specifically, the ImaRx Corporation (Tucson, Arizona) has developed thick fluid shelled agents, which release suspended taxol-based drugs from their shells upon destruction. Shape instabilities and surface waves correspond with the fragmentation and destruction of the agents. Finally, the interaction of multiple contrast agents has received little attention with respect to these emerging applications.

This paper examines a phenomenon suspected of limiting the spatial resolution of ultrasound elastography, and pursues a technical development to reduce the error introduced by this phenomenon. Elastography determines tissue strain by mapping post-motion image data onto pre-motion image data. An error arises in this mapping due to the fact that spatial warping of image data implies a warping of both the tissue and the measurement system response. An approach to data processing is studied to compensate for this system response related error. Underlying principals are first examined in one dimension, followed by presentation of at three dimensional implementation, suitable for application to volumetric ultrasound data.

This paper discusses two fabrication procedures used to build LiNbO₃ single element ultrasonic transducers with center frequencies in the 50-100 MHz range. Transducers of varying dimensions were built for an f-number range of 1.5- 3.0. A conductive quarter wavelength silver epoxy matching layer, and a conductive silver epoxy backing, were used in all designs. The desired focal depths were achieved by either casting an acoustic lens on the transducer face or press-focusing the piezoelectric into a spherical curvature. For lens-focusing transducers the lens material EPO-TEK 301 was modeled as a second matching layer. The pressed-focused transducer design utilized parylene as the second matching layer. For devices that required electrical impedance matching, a low impedance transmission line coaxial cable was used. All transducers were tested in a pulse-echo insertion loss arrangement, whereby the center frequency, bandwidth, insertion loss, and focal depth were measured. Several transducers were fabricated with center frequencies in the 50 to 100 MHz range. The measured -6dB bandwidths and two-way insertion loss values ranged from 33% to 70% and 12.5dB to 23.0dB, respectively. Both the lens-focusing and press-focusing techniques were successful in producing the desired focal depth without significantly compromising device sensitivity and bandwidth.

Ultrasound imaging at frequencies above 20 MHz relies almost exclusively on single-element transducers. IN order to apply array technology at these frequencies, several practical problems must be solved, including spatial scale and fabrication limitations, low device capacitance, and lack of a hardware beamformer. One method of circumventing these problems is to combine an array, an actuator, and a synthetic aperture software beamformer. The array can use relatively wide elements spaced on a coarse pitch. The actuator is used to move the array in short steps (less than the element pitch), and pulse-echo data is acquired at intermediate sample positions. The synthetic aperture beamformer reconstructs the image from the pulse-echo data. A 50 MHz example is analyzed in detail. Estimates of signal-to-noise reveal performance comparable to a standard phased array; furthermore, the actuated array requires half the number of elements, the elements are 8x wider, and only one channel is required. Simulated three-dimensional point spread functions demonstrate side lobe levels approaching - 40dB and main beam widths of 50 to 100 microns. A 50 MHz piezo-composite array design has been tested which displays experimental bandwidth of 70% while maintaining high sensitivity. Individual composite sub-elements are 18 microns wide. Once this array is integrated with a suitable actuator, it is anticipated that a tractable method of imaging with high frequency arrays will result.

A method for measuring the directivity function of transient fields with a new type of hydrophone that can be located at any convenient distance from the transducer is presented. Fields from planar and focused transducers, for both continuous wave and pulsed excitation, are measured via the new method, and the results compared against conventional measurements as well as against theoretical predictions. The directivity function for pulsed fields is best expressed as a complex directivity spectrum, and images of this fundamental transducer field characteristic are shown to encode a number of unexpected features. The definition and measurement of the directivity function, is not dependent on continuous wave or far-field conditions, and laboratory implementation of the theory is via a new type of hydrophone, with some unusual properties. It is concluded that precise and unambiguous measurement of transducer directivity patterns are straight forward to perform provided a relatively simple, but novel, technique is used. Images of the informative directivity spectrum may be obtained with ease.

Medical ultrasound research is typically performed using either video image data, or summed Radio Frequency (RF) data. While such data has led to improved understanding of ultrasound image formation, and in the development of novel image formation and signal processing algorithms, it contains only a fraction of the information available in the individual beamformer channels before summation. This paper describes the development of an advanced experimental system which will simultaneously acquire RF data from 128 individual beamformer channels. We refer to such data, acquired across the transducer face, as aperture domain data. The system will be capable of continuous acquisition over a period of 1.6 seconds, the equivalent of 50 image frames. The system will also incorporate a data interface to allow future connection to custom processing units, ultimately enabling real-time processing of aperture domain data. The system will be constructed around a state of the art Agilent Technologies SONOS 5500 ultrasonic imaging system to enable real-time imaging and preserve broad signal bandwidth, high signal to noise ratio, and high dynamic range. The proposed system will facilitate research on adaptive imaging, system architecture, multidimensional blood flow estimation, broadband transducers, and a number of other areas.

This paper provides results on processing a coherent bistatic ultrasonic array database that is collected at the University of Michigan. In this work, we use a scheme to synthesize the beam-steered and depth-focused array data of the target under study from its measured bistatic data. We also study the merits of processing these synthesized database via Fourier-based array imaging methods.

The feasibility of using a minimally-invasive pulse-echo diagnostic ultrasound technique for estimation of the spatiotemporal temperature distribution in the human prostate in vivo during transurethral microwave thermotherapy (TUMT) was investigated. The tissue temperature distribution was estimated using a method which allows spatially resolved measurement of changes in time of flight. Backscattered ultrasound RF signals were acquired from a commercially-available 7.5 MHz transrectal imaging probe and processed to compute and image the apparent tissue internal displacement distribution. The temperature distribution was determined by exploiting an estimated initial sound speed distribution and an empirical relationship between temperature and ultrasound transmission speed. Results form an in-vitro phantom study showed a linear relationship between induced displacement and temperature for a limited temperature range. Initial results from an exploratory in-vivo study in patients undergoing TUMT showed substantial influence form tissue motion and deformation. Problems inherent to in-vivo data acquisition are discussed and potential solutions are proposed. Possible approaches to further improve the algorithm and to assess the ability of the system to estimate and monitor the intraprostatic temperature distribution during TUMT are presented.

A C-scan through-transmission ultrasound system has been constructed based on a patented hybrid microelectronic array that is capable of generating ultrasound images with fluoroscopic presentation. To generate real-time images, ultrasound is introduced into the object under study with a large unfocused plane wave source. The resultant pressure wave strikes the object and is attenuated and scattered. The device detects scattered as well as attenuated ultrasound energy which allows the use of an acoustic lens to focus on detected energy from an object plane. The acoustic lens collects the transmitted energy and focuses it onto the ultrasound sensitive array. The array is made up to two components, a silicon detector/readout array and a piezoelectric material that is deposited onto the array through semiconductor processing. The array is 1 cm on a side consisting of 128x128 pixel elements with 85micrometers pixel spacing. The energy that strikes the piezoelectric material is converted to an analog voltage that is digitized and processed by low cost commercial video electronics. The images generated by the device appear with no speckle artifact with fluoroscopy-like presentation. The images show no obvious geometrical distortion. The experimental results indicated that the system has a spatial resolution of 0.32 mm. It can resolve 3mm objects with low differential contrast and an attenuation coefficient difference less than 0.07 dB/cm/MHz. Phase contrast of the objects are also clearly measurable. A presentation of a C- scan image guided breast biopsy was demonstrated. In addition, punctured needle tracks in a tumor was clearly observed. This implies the potential of observing the spiculation of masses in vivo.

Waveform synthesis for pulse-echo medical ultrasonic imaging, in conjunction with post-beamforming filtering, will undoubtedly play an important role in defining the ultimate quality of the next generation of medical ultrasonic imaging. Two important applications that will rely heavily on appropriate waveform synthesis are contrast- assisted imaging and multi-modal high-speed imaging with parallel processing of multiple image lines using coded excitation and filterbank reconstruction. In this paper, we address the issue of beam-space waveform synthesis using ultrasound phased arrays typically used in medical pulse- echo imaging applications. Simulation and experimental results will be presented to illustrate the role of the transducer aperture and bandwidth characteristics in the waveform synthesis. Furthermore, we generalize the concept of point-spread function (PSF) to allow for the post- beamforming filter-based image reconstruction. We show experimentally that these PSFs serve as a reliable predictor of the image quality for multi-modal pulse-echo imaging systems employing post-beamforming filter-based reconstruction. Combined with a computationally efficient Fourier-based method for their derivation, these PSFs thus serve as a powerful tool in the design and optimization of coded waveforms for pulse-echo imaging applications. A description of the waveform synthesis algorithm will be given along with illustrative examples using computer simulations and experimental results from tissue-mimicking phantoms.

We are investigating a novel ultrasonic method for remote palpation, which provides images of local variations in tissue stiffness. Acoustic radiation force is applied to small volumes of tissue, and the resulting displacement patterns are imaged using ultrasonic correlation based techniques. Tissue displacements are inversely proportional to tissue stiffness, thus a stiffer region of tissue exhibits smaller displacements than a more compliant region. This method also provides information about tissue recovery after force cessation. We will present in vivo experimental results demonstrating the feasability of this method. Using intensities ranging from 120 to 300 W/cm², peak displacements of up to 50 microns were observed after 1.4 milliseconds of force application. The tissue moved to its peak displacement within 3 milliseconds of force application, and the time constants for tissue recovery varied with tissue type. Tissue displacements appeared to be correlated with tissue structure in matched B-mode images. To our knowledge, these results represent the first in vivo soft tissue images generated using radiation force. These findings support the feasibility of Remote Palpation imaging. We will discuss the technical, safety, and clinical challenges of implementing a real-time Remote Palpation imaging system on a commercial diagnostic scanner.

Time-resolved 2D Pulsed Elastography is a new elastographic technique for imaging the shear modulus of soft tissues. A low-frequency transient shear wave is sent in the medium while an ultra-fast ultrasonic imaging system acquires frames at a very high frame rate (up to 10,000 frames/s). This ultra-fast ultrasonic imaging system has been specifically developed for this application. It is based on a time-reversal mirror of 128 channels sampled at 50 MHz and having 2 Mbytes random access memory. Displacements induced by the slowly propagating shear wave are measured using the standard cross-correlation technique. The low-frequency excitation is obtained with a device composed of two rods that are placed around the ultrasonic transducer linear array. The rods vibrate perpendicularly to the surface of the tissues. They may be either parallel or perpendicular to the active surface of the array. With this device, large amplitude displacements are observed in the ultrasonic image area. We have measured the spatio- temporal evolution of the displacements induced by the low- frequency (20-100 Hz) shear wave in tissue-equivalent phantoms and breast in-vivo. A direct local inversion is used to recover the shear modulus distribution map in phantoms containing hard regions.

In Ultrasound-Stimulated Vibro-Acoustography (USVA) imaging method, one tries to form an image of the deformability of a tissue submitted to a low frequency (LF) f_- stress field. This sound field can be locally created by mean of a focused annular array emitting two primary beams driven at the two close frequencies f₁ and f₂ = f$=1)+f_-. The coherent acoustic emission resulting form the object vibration is detected by a sensitive hydrophone and used to form an image. In the present literature, the origin of this stress field has been essentially identified to the LF radiation pressure created by the two primary beams frequency beating. However, an other contribution of this internal stress is, according to us, the LF field distributed in the object volume and created by the nonlinear (NL) interaction of the two primary beams. The q_-, q₁ and q₂ beams emitted by a focused four rings annular array are experimentally measured in a water tank. Amplitude and shape of the theoretical model of the NL interference beam and the experimental curves are compared. USVA images of a steel ball placed into gelatin and scans of a calcaneus bone at different depths are presented. The resolution of an elastodynamic problem in the case of a spherical element in gelatin shows that the displacement amplitude of the tissue, in force axis, is principally due to the variations of the elasticity shear modulus (mu) . USVA images represent in gray scale the displacement amplitude due to the radiation resulting from the object vibration. If we assume that the stress field is constant during the scan, we then obtain images proportional to the local elasticity shear modulus.

This report summarizes our experiments with novel ultrasound phantoms that mimic essential bio mechanical and dynamical vascular features of soft biological tissues. Real-time RF echo acquisition using a Siemens Elegra ultrasound system at 7.5 MHz provided a time series of ultrasonic images that were used to image longitudinal strain from pulsed and steady flows. Physical features of internal deformation patterns resulting from pulsatile flows revealed that ultrasonic strain imaging could be a very sensitive method for observing important properties related to physiological fluid flow.

A technique called disparity mapping (DM) processes pairs of ultrasound B-scan images collected while a sonographer varies the probe pressure slightly on the breast surface. Dm measures the apparent displacement of the tissue about each image point and subsequently constructs a correlation map which represents the similarity between the speckle patterns around each point. The continuity of the lesion perimeter in the correlation image is used to separate benign from malignant lesions, with high continuity corresponding well with benign lesions and highly segmented perimeters correlating with malignancies. Twenty five solid masses were evaluated, and the results were compared with histology from core or surgical biopsy, or with cytology from fine needle aspiration. The results analyzed all lesions correctly (15 cancers and 10 benign lesions). There were no false positives or false negatives. The results suggest that DM may be a useful tool in digitally diagnosing breast lesions and consequently in reducing the number of unnecessary biopsies.

The correlation between ultrasonic speckle patterns acquired with different system configurations determines the effectiveness of spatial and frequency compounding, phase aberration correction, elastography, and other applications. For some algorithms, such as compounding, decorrelation has a direct effect on performance. In others decorrelation indirectly degrades performance by reducing the accuracy of signal processing methods such as time delay estimation (TDE). This paper reviews the causes of echo decorrelation for a variety of applications. The mechanisms of decorrelation are described using k-space and simulation. We show examples of signals at different correlation levels to build intuition. The direct impact of decorrelation is described for several of the aforementioned applications. We quantify the indirect effect of decorrelation by examining TDE performance. This analysis shows that reducing correlation from 1.0 to 0.99 increases TDE errors by a factor of 3 under reasonable conditions. This paper also describes a new method of computing speckle pattern correlation directly from system point spread functions, assuming either shift- invariant or shift-variant systems. This paper concludes that echo decorrelation has a major impact on the performance of ultrasound signal processing, and that this decorrelation can be computed using simulation, k-space, or the system point spread function.

We acquired conventional and harmonic channel r.f. ultrasound echo data using a 3x80 element, 8.5 MHz multirow array from the breasts of eight volunteers. The data acquisition was interleaved to allow direct comparison normal and harmonic echo wavefronts. Harmonic imaging data was acquired using the pulse inversion technique. Data was acquired form extended regions of interest (25 mm deep, 10 mm wide). Time shift estimates from pairs of elements were combined using a weighted least squares algorithm to obtain a wavefront arrival time error estimate. Low spatial frequencies dominated most of the wavefront estimates, and many had a curvature suggesting a gross sound speed error. Wavefront estimates were often stable over lateral translations of a few millimeters, although they often changed significantly with range, particularly at tissue boundaries observed in the B-mode image. Averaging wavefront estimates over range yielded phase aberration estimates that generally improved image quality. We measured relatively small wavefront arrival errors with both conventional (22.9 +/- 7.6 ns r.m.s.) And harmonic (22.8 +/- 8.8 ns r.m.s.) echoes. For any particular measurement, the difference between conventional and harmonic wavefront estimates was small (0 +/- 4.5 ns r.m.s.). Our measurements suggest relatively mild phase aberrations in the breast, although they may be more significant for higher frequency transducers and deeper imaging depths.

Ultrasonic imaging systems capabilities are strongly dependent on the focusing quality of the ultrasonic beam. In the case of brain imaging, the skull strongly degrades the ultrasonic focusing pattern by introducing strong phase and amplitude aberrations of the wavefront. In previous work, this degradation of the beam focus has been partially corrected by coupling the time reversal focusing process to an amplitude compensation of the emitted signals. In that case, the optimal focus was reproduced down to -20 dB, but the sidelobe level remained at about -25 dB. We propose here a new focusing technique, called spatio- temporal inverse filter, based on the inversion of the propagation operator at each frequency within the bandwidth of our transducers. Simulations and experiments will be presented that show the effectiveness of the technique. Experimental focusing through the skull is now comparable to the optimal focusing in homogeneous medium. It will be explained how the set of emission signals can be used in order to achieve both transmit and receive focusing. In the transmit-receive mode, focusing through the skull reaches the optimal level obtained in water down to -70 dB (i.e. constrained only by experimental noise levels). This could lead to high quality real time brain imaging and Doppler flow mapping.

Low contrast detectability of diagnostic ultrasound is fundamentally limited by speckle brightness variations. Several compounding techniques have been proposed for speckle reduction by incoherently averaging partially correlated measurements. The conventional approaches obtain partially correlated measurements by imaging the same object from different spatial positions or within different frequency ranges. In this paper, a new technique based on signal decorrelation under different strain states is proposed. The different strain states can be created using external applied forces. By correcting on ly the in-plane motion, images under different strain states have identical characteristics except for speckle appearance due to the un- corrected out-of-plane motion. Therefore, these images can be combined for speckle reduction without significantly affecting the in-plane spatial resolution. Efficacy of the new compounding technique was previously tested using data acquired by a single crystal transducer. However, only the axial motion was corrected and RF data were used to estimate the displacement fields. In this paper, results are extended to full two-dimensional motion correction using post-detection images from a commercial array system. Images from human thyroids and gelatin based phantoms are investigated. It is shown that speckle brightness variation's can be effectively reduced without significant degradation in spatial resolution.

Since the development of the directional Doppler by McLeod in 1967, methods of acquiring, analyzing, and displaying blood velocity information have been under constant exploration. These efforts are motivated by a variety of interest and objectives including, to: a) simplify clinical examination, examiner training, and study interpretation, b) provide more hemodynamic information, and c) reduce examination variability and improve accuracy. The vector Doppler technique has been proposed as one potential avenue to achieve these objects. Vector Doppler systems are those that determine the true 2D or 3D blood flow velocity by combining multiple independent velocity component measurements. Most instruments can be divided into two broad categories: 1) cross-beam and 2) time-domain. This paper provides a brief synopsis of the progression of vector Doppler techniques, from its onset in 1970 to present, as well as possible avenues for future work. This is not intended to be a comprehensive review of all vector Doppler systems.

Ultrasound vector flow estimation ben be complicated by the spatial anisotropy of the ultrasound pulse, a limitation most commonly characterized in terms of the different spatial resolutions achieved along the three axes of a typical imaging system. For vector Doppler or spatial quadrature flow estimation techniques, this anisotropy limits, relative to the axial modulation, the maximum achievable frequency of the modulation associated with lateral motion. This anisotropy has one advantage in establishing higher aliasing thresholds in the non-axial dimensions than are typically encountered axially. However, it is likely that one of the obstacles to practical implementation of vector velocity estimation will be the attenuation of non-axial modulation components by conventional wall and clutter filters. As part of our investigation of this topic, we describe a commonality among several vector flow estimators: the vector Dopple described by Overbeck, et. al, the estimator described by Jensen and Munk, and the heterodyning spatial quadrature estimator we have previously described. When implemented with the same transducer array, these estimators produce very similar spatial impulse responses despite the use of radically different complex apodization schemes. In two cases this common response is not formed during beamforming but rather after post-processing has been applied. We describe this isomorphism and discuss the differences among the estimators that we expect to become evident under realistic imaging conditions. We present an analysis of the multi-dimensional spatial frequency responses of these vector velocity estimators in comparison with typical wall filter responses. The axial/lateral anisotropy for all three estimators was found to be on the order of 10 in a pulse-echo regime. This implies that detection of lateral flow at a rate of, for example, 10 m/s would require a wall filter setting providing an axial cut-off velocity of less than 1 m/s. The practical implications of these findings and alternative approaches are discussed.

Estimation of velocity vectors using transverse spatial modulation has previously been presented. Initially, the velocity estimation was improved using an approximated dynamic beamformer setup instead of a static combined with a new velocity estimation scheme. A new beamformer setup for dynamic control of the acoustic field, based on the Pulsed Plane Wave Decomposition (PPWD), is presented. The PPWD gives an unambiguous relation between a given acoustic field and the time functions needed on an array transducer for transmission. Applying this method for the receive beamformation results in a setup of the beamformer with different filters for each channel for each estimation depth. The method of the PPWD is illustrated by analytical expressions of the decomposed acoustic field and these results are used for simulation. Results of velocity estimates using the new setup are given on the basis of simulated and experimental data. The simulation setup is an attempt to approximate the situation present when performing a scanning of the carotid artery with a linear array. Measurement of the flow perpendicular to the emission direction is possible using the approach of transverse spatial modulation. This is most often the case in a scanning of the carotid artery, where the situation is handled by an angled Doppler setup in the present ultrasound scanners. The modulation period of 2 mm is controlled for a range of 20-40 mm which covers the typical range of the carotid artery. A 6 MHz array on a 128-channel system is simulated. The flow setup in the simulation is based on a vessel with a parabolic flow profile for a 60 and 90-degree flow angle. The experimental results are based on the backscattered signal from a sponge mounted in a stepping device. The bias and std. Dev. Of the velocity estimate are calculated for four different flow angles (50,60,75 and 90 degrees). The velocity vector is calculated using the improved 2D estimation approach at a range of depths.

The blood velocity can be estimated by cross-correlation of received RF signals, but only the velocity component along the beam direction is found. A previous paper showed that the complete velocity vector can be estimated, if received signals are focused along lines parallel to the direction of the flow. Here a weakly focused transmit field was used along with a simple delay-sum beamformer. A modified method for performing the focusing by employing a special calculation of the delays is introduced, so that a focused emission can be used. The velocity estimation was studied through extensive simulations with Field II. A 64-elements, 5 MHz linear array was used. A parabolic velocity profile with a peak velocity of 0.5 m/s was considered for different angles between the flow and the ultrasound beam and for different emit foci. At 60 degrees the relative standard deviation was 0.58% for a transmit focus at 40 mm. For 90 degrees the new approach gave a relative standard deviation of 8.3% with a focus at 40 mm and 8.0% at a transmit focus of 150 mm. Pulsatile flow in the femoral artery was also simulated. A purely transverse flow profile could be obtained with a relative standard deviation of less than 10% over the whole cardiac cycle, which is sufficient to show clinically relevant transverse color flow images.

The aim of the vector Doppler (VD) technique is the quantitative reconstruction of a velocity field independently of the ultrasonic probe axis to flow angle. In particular vector Doppler is interesting for studying vascular pathologies related to complex blood flow conditions. Clinical applications require a real-time operating mode and the capability to perform Doppler measurements over a defined volume. The combination of these two characteristics produces a real-time vector velocity map. In previous works the authors investigated the theory of pulsed wave (PW) vector Doppler and developed an experimental system capable of producing off-line 3D vector velocity maps. Afterwards, for producing dynamic velocity vector maps, we realized a new 2D vector Doppler system based on a modified commercial echograph. The measurement and presentation of a vector velocity field requires a correct spatial sampling that must satisfy the Shannon criterion. In this work we tackled this problem, establishing a relationship between sampling steps and scanning system characteristics. Another problem posed by the vector Doppler technique is the data representation in real-time that should be easy to interpret for the physician. With this in mine we attempted a multimedia solution that uses both interpolated images and sound to represent the information of the measured vector velocity map. These presentation techniques were experimented for real-time scanning on flow phantoms and preliminary measurements in vivo on a human carotid artery.

The tortuous geometry of the human vasculature system, and pulsatile nature of blood flow, combine to generate a complex 3D blood flow pattern. Conventional Doppler ultrasound, however, provides velocity data only in one dimension - in the direction of the ultrasound beam. 3D velocity results are extrapolated form this one dimensional data by assuming the blood is parallel to the vessel axis. A two-dimensional vector Doppler system has been developed to improve hemodynamic visualization, and investigate the possible errors introduced by conventional Doppler methods. The results from a study of a normal human femoral bifurcation demonstrate that flow is not always para-axial. When flow is para-axial, conventional Doppler methods work as expected: the calculated velocity is independent of interrogation angle. When flow is non-axial, however, the velocity waveforms and measurements vary greatly with interrogation angle.

The first high frequency ultrasound system able to image blood flow in the microcirculation in real-time has been developed. 2D color flow frames are rapidly acquired using a recently reported method to achieve frame rates approaching 10 fps. A new flow phantom was constructed in order to tune the wall filter order, cutoff and attenuation for a 25 MHz, f/2 transducer. RF data were acquired in both M-mode and swept-mode, and processed in order to tune the wall filter. These filters were then used in making controlled measurements of flow velocity and volume flow rate for a typical PRF of 500 Hz (1 mm/sec scan speed). Over th einput of mean axial velocities ranging from 0.3 to 3.0 mm/sec (0.88 to 8.8 mm/sec angle corrected), the measured mean and maximum flow velocities were linear, with slight over-estimation of mean velocities due to the wall filter cutoff. Without correction for finite beam size, the volume flow rates were over-estimated by a factor of 2. The color flow settings were then applied to image microcirculatory flow within the nail bed of a human finger, where they were tested and optimized for a variety of vessel sizes and flow velocities.

The effect of signal decorrelation on the performance of the Butterfly Search velocity estimator is examined. An analytical approximation for the expected value of the Butterfly Search L(v) function is developed for three cases of interest. The approximations are verified against synthesized echo data. It is found that the peak value of the L(v) function is limited by the rate of signal decorrelation. The results show that improved performance may be obtained by processing and averaging subsets of echo ensembles, rather than applying the Butterfly Search to the entire ensemble simultaneously. For lower SNRs, processing the entire ensemble at once produces equivalent or better results than subset processing. Results from echo data obtained in-vitro are presented which confirm the simulations.

A unique in-vitro system has been developed that incorporates both realistic phantoms and flow. The anthropomorphic carotid phantoms are fabricated in agar with stenosis severity of 30% or 70% (by NASCET standards) and one of two geometric configurations- concentric or eccentric. The phantoms are perfused with a flow waveform that simulates normal common carotid flow. Pulsed Doppler ultrasound data are acquired at a 1 mm grid spacing throughout the lumen of the carotid bifurcation. To obtain a half-lumen volume, symmetric about the mid plane, requires a 13 hour acquisition over 3238 interrogation sites, producing 5.6 Gbytes of data. The spectral analysis produces estimates of parameters such as the peak velocity, mean velocity, spectral-broadening index, and turbulence intensity. Color-encoded or grayscale-encoded maps of these spectral parameters show distinctly different flow patterns resulting from stenoses of equal severity but different eccentricity. The most noticeable differences are seen in the volumes of the recirculation zones and the paths of the high-velocity jets. Elevated levels of turbulence intensity are also seen distal to the stenosis in the 70%-stenosed models.

In conventional elastography, strains are estimated by computing gradient of estimated displacement. However, gradient-based algorithms are susceptible to noise. We have developed two new strain estimators to overcome the common limitations of elastography. The first estimator is based on a frequency-domain formulation; it estimates local strain by maximizing the correlation between the spectra of pre- and post-compression echo signals by iteratively frequency- scaling the latter. We discuss a variation of this algorithm that may be computationally more efficient. The second estimator is based on the observation that an extremely stiff region will undergo virtually no strain when compressed, and will exhibit quasi-rigid body motion. As a result, an area with high similarity between the pre- and post-compression signals indicates low strain, and an area with low similarity indicates large strain. We use normalized 2D correlation function to estimate this similarity. This method offers significant advantages for detecting rigid tissues in the presence of large, irregular, non-axial motion. Both the estimators exhibited promising results in simulation and experiments.

During neurosurgery, intraoperative brain shift comprises the accuracy of image guided techniques. We are investigating the use of ultrasound as an inexpensive means of gaining 3D data on subsurface tissue deformation. Measured displacement of easily recognizable features can then be used to drive a computational model for a description of full volume deformation. Subsurface features identified in the ultrasound image plane are located in world space using a 3D optical tracking system mounted to the ultrasound scanhead. This tracking system is also co- registered with the model space derived from preoperative MR, allowing the ultrasound image plane to e reconstructed in MR space, and the corresponding oblique MR slice to be obtained. The ultrasound image tracker has been calibrated with a novel strategy involving multiple scans of N-shaped wires positioned at several depths. Mean calibration error is found to range from 0.43 mm to 0.76 mm in plane and 0.86 mm to 1.51 mm out of plane for the two ultrasound image scales calibrated. Improved ultrasound calibration and co- registration facilitates subsurface feature tracking as a first step in obtaining model constraints for intraoperative image compensation. Estimation of and compensation for brain shift through the low cost, efficient technology of ultrasound, combined with computational modeling is feasible and appears to be a promising means of improving intraoperative image guided techniques.

In quantitative tissue characterization. Obtaining processed ultrasonic echoes with a direct relationship to local tissue response (backscatter spectrum) and that are free from systemic depth-dependent effects, such as diffraction, is essential. In general practice today, these unwanted distortions are eliminated by dividing short time power spectra. However, this method has its drawbacks; noise is not taken into account, and shorter time gates lead to an increasing bias within the relative spectra. To overcome these methodological issues, I propose a different approach as follows. Entire deconvolved A-scans are estimated by a Kalman smoothing deconvolution algorithm. These then serve as a basis for estimating the relative backscatter spectra. In addition, due to the principle of the deconvolution algorithm, it is possible to suppress additive noise to some degree. To examine the properties of the method proposed, this paper presents an analytical expression for the power spectrum of the deconvolved signals obtained by Kalman Smoothing. This result is then compared to the expectations of relative short time power spectra. Simulations demonstrate the behavior of the deconvolution method in a non-stationary environment.

Purpose: The cornea acts as the window of the eye's optical system, and its chief refractive component. The lack of intervening tissues makes the cornea accessible to very high frequency ultrasonic study. In this report, we detail use of radiofrequency (RF) signal processing methods to characterize corneal pathology and to enhance biometric precision. Methods: Using a 50 MHz PVDF transducer, we scanned the cornea using an arc motion so as to maintain normality and consistent range. RF data were acquired at a sample rate of 500 MHz. Deconvolution against a glass-plate echo allowed biometric enhancement (by effectively broadening the bandwidth) and measurement of tissue backscatter spectra. Results: Calibrated spectrum analysis was used to quantitatively measure backscatter in corneal scars and other pathologies. Signal processing allowed us to attain reproducibility for repeated measurements of the corneal epithelium (approximately 50 microns thick) to 1 micron. By combining measurements from a series of planes, maps of the thickness of the individual layers comprising the cornea were produced. Conclusion: The layers of the cornea have different optical refractive indices, and thus their thicknesses directly affect visual acuity. The scattering of light by a corneal scar is caused by inhomogeneities or irregularities that may result in acoustic backscatter as well. The ability of ultrasound to quantify backscatter and corneal layer thickness provides a new avenue for diagnosis of corneal disease and refractive abnormalities.

Many studies have demonstrated that time-domain speed-of- sound (SOS) measurement sin calcaneus are predictive of osteoporotic fracture risk. However there is a lack of standardization for this measurement. Consequently, different investigators using different measurement systems and analysis algorithms obtain disparate quantitative values for calcaneal SOS, impairing and often precluding meaningful comparison and/or pooling of measurements. A numerical method has been developed to model the effects of frequency- dependent attenuation and dispersion on transit-time-based SOS estimates. The numerical technique is based on a previously developed linear system analytic model for Gaussian pulses propagating through linearly attenuating, weakly dispersive media. The numerical approach is somewhat more general in that it can be used to predict the effects of arbitrary pulse shapes and dispersion relationships. The numerical technique however utilizes several additional assumptions (compared with the analytic model) which would be required for the practical task of correcting existing clinical databases. These include a single dispersion relationship for all calcaneus samples, a simple linear model relating phase velocity to broadband ultrasonic attenuation, and a constant calcaneal thickness. Measurements on a polycarbonate plate and thirty human calcaneus samples were in good quantitative agreement with numerical predictions. In addition, the numerical approach predicts that in cancellous bone, frequency-dependent attenuation tends to be a greater contributor to variations in transit-time-based SOS estimates than dispersion. This approach may be used to adjust previously acquired individual measurements so that SOS data recorded with different devices using different algorithms may be compared in a meaningful fashion.

Frequency dependent attenuation is a pronounced feature of ultrasound propagation in human tissues. A new technique for calculating that parameter from backscattered echoes is demonstrated, and it is shown how the information may be incorporated into a gray scale image. Analysis of the noisy nature of attenuation estimates from conventional backscattered echo signals suggests a new technique for producing less erratic estimates, by manipulating zeroes in the complex frequency domain. No signal averaging is needed, and the method lends itself to analysis of short data segments, thereby providing suitable input for attenuation imaging. Rf data is required. The new method is found to considerably reduce the variance of the attenuation estimates from short data segments. It is found that attenuation-weighted B-mode images are an informative way to show results. The method of zero manipulation, presented here, for producing less noisy pulse-echo attenuation estimates represents a powerful approach towards the problem.

Conventional B-mode ultrasound is the standard means of imaging the prostate for guiding prostate biopsies and planning radiotherapy (i.e., brachytherapy and external-beam radiation) of prostate cancer (CaP). Yet B-mode images essentially do not allow visualization of cancerous lesions of the prostate. Ultrasonic tissue-typing imaging based on spectrum analysis of radio-frequency (RF) echo signals has shown promise for overcoming the limitations of B-mode imaging in distinguishing cancerous from common forms of non-cancerous prostate tissue. Such tissue typing utilizes non-linear methods, such as nearest-neighbor and neural- network techniques, to classify tissues based on spectral- parameter and clinical-variable values. Our research seeks to develop imaging techniques based on these methods for the purpose of improving the guidance of prostate biopsies and the targeting of brachytherapy and external-beam radiotherapy of prostate cancer. Images based on these methods have been imported into real-time instrumentation for biopsy guidance and into commercial dose-planning software for real-time brachytherapy. 3D renderings show locations and volumes of cancer foci. These methods offer exciting possibilities for effective low-cost depiction of prostate cancer in real time and off-line images. Real-time imaging showing cancerous regions of the prostate can be of value in directing biopsies, determining whether biopsy is warranted, assisting in clinical staging, targeting brachytherapy, planning conformal external-beam radiation procedures, and monitoring treatment.

In medical ultrasonic imaging the signal reflected from the tissue often has a random character to it. It is believed that the random nature of the tissue scattering microstructure is responsible for the stochastic nature of the echo signal. Chen, et. al. Have proposed a signal processing scheme that is based on the statistical moments calculated on the Fourier transform of the time gated echo signal. The theory requires the knowledge of a frequency- dependent effective cell volume term. This paper describes the use of a closed form expression (Lommel diffraction formulation) for this purpose. Our simulation results suggest that reliable estimation of the cell volume is possible only when the time duration of the excitation pulse is small compared to the time gate length.

The objective of this study is to investigate the feasibility of elastographic monitoring of High Intensity Focused Ultrasound (HIFU) therapy of prostate cancer. Elastography is an imaging technique based on strain estimation in soft tissues under quasi-static compression. Since pathological tissues and HIFU-induced lesions exhibit different elastic properties than normal tissues, elastography is potentially able to achieve these goals. An ultrasound scanner was connected to a PC to acquire RF images. This setup is compatible with a HIFU device used for prostate cancer therapy by transrectal route. The therapy transducer and the biplane-imaging probe are covered with a balloon filled with a coupling liquid. Compression of the prostate is applied by inflating the balloon, while imaging sector scans of the prostate. In-vivo elastograms of the prostate were acquired before HIFU treatment. Problems inherent to in-vivo acquisitions are reported, such as undesired tangential displacements during the radial compression. This study shows the potential for in-vivo elastogram acquisition of HIFU-induced lesions in the human prostate.

This paper presents an algorithm that extracts accurate left ventricular (LV) boundaries from a 2D echocardiographic (echo) sequence covering a cardiac cycle. Unlike user- dependent, manual or semi-automatic techniques, the key feature of this algorithm is its truly automated processing for estimation. First, the algorithm performs smoothing of the image in the LV target area, followed by enhancement of intensity differences and edge detection. In order to best localize the position of the LV boundary, the algorithm uses a deformable template model derived from prior knowledge of LV shape and an edge map obtained from boundary estimation. The deformable template model is matched to the target by minimizing an energy function induced by the difference between the edge locations and tangents of the template and those of the current frame edge map. Since the shape of the endocardial boundary will vary between temporally distinct frames, a controlled continuity spline, a snake, is then used to implement refined active contour matching to the current frame LV boundary. Frame-to-frame tracking of the LV boundary is incorporated by using the boundary estimate from one frame to initialize and help with the estimation in the subsequent frame, which leads to faster and more accurate LV estimation throughout the image sequence. Test results of this algorithm show that the combination of approximate template matching with smoothness constraints in snakes produces good LV boundary extraction even with significant false and/or missing edge information caused by poor contrast and noise.

It is very important to observe the vessels of the patient who are dialyzed artificially. An X-ray examination using contrast medium injected to the patient has been used for this purpose up to the present, but sometimes the examination has a risk of radiation damage. Therefore, we developed a safe and easy-to-use system in which 3D images of the vessels in the patients are reconstructed very quick from ultrasonic echoes. In this system, a view point for 3D rendering is set on the above position of the ultrasonic transducer, and a ray for the rendering is coincided with an ultrasonic beam. These features enable 3D images to be gradually reconstructed in real time while the echoes are being received. A magnetic position sensor system and a special 3D scanner which was developed were adopted for acquiring 3D echo data. In signal processing, intensity inversion technology is carried out before the 3D rendering process in order to detect and emphasize the vessels. With this system, we have acquired echo signals from the vessels in the arm of kidney dialyzed patients and made similar 3D images of X-ray angiography with the echoes in a short time such as 4 to 8 seconds.

We propose an efficient method for content-based ultrasound image retrieval using magnitude frequency spectrum and implement an ultrasound image retrieval system based on the proposed method. The target images are ultrasound images of adult organs. A trained staff often acquires such images so that images of the same kind of organs are very similar, although their locations may not exactly coincide. Therefore, the magnitude frequency spectrum, which has a translation-invariant property, is used as a feature for content-based retrieval. A test image database is composed of real ultrasound images. As a retrieval result, a specified number of highly similar target images are retrieved from all the target images. If all the target images in the database are pre-classified into organs of the same kind, the retrieved images are selected among the images whose class is the same as that of the highest similarity image. Experimental results of the proposed method is superior to other methods. The proposed method especially yields further performance improvement by using the pre-classification. Moreover, it is found from the experimental results that the magnitude frequency spectrum method is robust to the speckle noise that usually exists in ultrasound images.

Although broadband ultrasonic attenuation from calcaneus has been demonstrated to be useful in the diagnosis of osteoporosis, the processes governing the interactions between ultrasound and bone are cunently not well understood. Attenuation is the combined result of two components: absorption and scattering. The objectives of this study were to 1) develop and test a theoretical model for scattering from calcaneus, and 2) investigate the relative roles of absorption and scattering in determining attenuation.

An exposimetry system for characterization of high frequency ultrasound fields has been developed and built. By extrapolating the recommendations of the AIUM and IEC standards to higher frequencies, an exposimetry system operating above 15 MHz was outlined. The system incorporates a five degrees-of-freedom positioning system, including three automated translational motors that provide 0.5 micron resolution. Two manual rotational axes utilize a worm-gear and concentric cylinder arrangement to insure orthogonal rotational adjustment. Overall bandwidth of the system is 100 MHz and is limited by the type of hydrophone used. Using a calibrated 0.04 mm diameter needle-type hydrophone, measurements of single element transducers of 25-50 MHz have been made. LiNbO₃ and PVDF transducers of f-numbers from 2-3 have been tested and 2D intensity beam profiles plotted. Results from a 50 MHz LiNbO₃ transducer show good agreement between empirical (8.6 mm) and theoretical (9.0 mm) focal points. The -3 dB beamwidth was also measured (108 micron) to be comparable to that of the calculated value (86 micron). It is shown that this system provides a good means for characterization and analysis of the beam profiles of high frequency transducers.

The novel solution of the KZK equation for acoustic pressure of the second harmonic in slightly focused beam of a circular transducer was obtained in a closed form for moderately nonlinear absorbing media (Gol'dberg numbers ~ 1). The solution is based on the method of slowly changing wave profile in combination with the method of successive approximations. Two pairs of transducers (Valpey-Fisher Corp.) Were compared to investigate the influence of focusing on the applicability of the moderate nonlinearity approach. The first pair was of 0.25' diameter and the second was of 0.5' diameter. Both pairs has one transducer with flat surface and the other geometrically focused at 4'. The central frequency for all transducers was 5 MHz. Measurements were undertaken in the blood-mimicking solution of water and glycerine. The results demonstrated that for slightly focused transducers with circular apertures, the moderate nonlinearity approach is still valid, as it was proved for flat sources with the same source level, despite the higher pressures in the focal region. The peak pressure for the weakly focused system occurs at a shorter range than focal length.

We propose two techniques for robust motion artifact suppression in ultrasound images based on what is called navigator echoes. The navigator echo approach has been introduced and widely used in the area of magnetic resonance imaging where motion is detected along one of the image directions based on simple registration of one-dimensional projection of the image assuming rigid body motion. In ultrasound imaging, the same concept can be applied to correct for simple shifts (e.g., with linear array probes) or rotations (e.g., with convex array probes) along the width and depth direction. The first technique assumes a rigid body motion model. Hence, simple one-dimensional template matching can be used to obtain the motion parameters as in magnetic resonance imaging. Alternatively, the second technique considers a more general spatially- variant motion model. The motion parameters of this model can be obtained through localized optimization of an information theoretic criterion. The motion model parameters are then employed in the reconstruction thus providing images with substantially reduced artifacts. The proposed method was implemented on an experimental ultrasound system in which each line is obtained from 2-4 acquisitions from interleaved frames. The motion models are compared and their practical implementation for clinical systems is evaluated.

In speckle motion tracking, blood velocity magnitude and direction are estimated from speckle pattern changes between successive images based on either 2D correlation of sum- absolute-difference (SAD) methods. Even though these techniques have been proven useful for flow mapping applications, they suffer from bias effects in estimating due to the presence of clutter induced from structural motion. In this work, we propose a technique for reducing the clutter effect, and hence enhancing the robustness of velocity estimation. The proposed technique relies on separating the speckle from the underlying specular structures. The basic idea is to employ a speckle reduction strategy based on nonlinear coherent diffusion filtering to obtain speckle free image of vessels from an original B- mode. Then, subtracting such image from the original image, an image for speckle is obtained. Nonlinear coherent diffusion filtering has been proven successful in removing Rayleigh distributed speckle pattern resulting mainly from blood scatterers in B-mode images while preserving structural information. This allows such scattering pattern to be utilized more accurately in calculating the velocity using an ultrasound research system and velocity estimates were obtained using 2D correlation.

Several methods exploit the relative motion between the probe and the object being scanned to figure out an estimate of the normals of the existing structures in a volume. These methods are revealed as a good estimator for normals, at least better than simple gradient schemes. On the other hand, polygonal meshes can be obtained directly from raw data by means of tiling algorithms. Although these meshes are good representations of isosurfaces in CT or MRI data, as far as ultrasound is concerned, results are quite noisy, so more effort is needed in developing algorithms that will be able to enhance the structures in the images. In this paper we propose a method that reshapes the geometry of meshes using the information given by normals. Rendering the meshes with the estimated normals is meaningful smoothness is observed. Therefore it is reasonable to obtain a new geometry for the meshes by imposing the normals as an external condition. In order to achieve coherence between the two entities (polygonal meshes and normals), a local optimization approach is proposed. For each vertex, the position that minimizes the norm of the error between the geometric normal and the external normal is worked out. A second term in the objective function favors solutions that are closer to the current state of the mesh. This minimization process is applied to all vertices that constitute the mesh and it is iterated so as to find a global minimum in the objective function. Our results show a better match of external normals and meshes, which draws more natural surface-rendered images.

We have designed and evaluated (using computer simulations) several different 2D phased arrays with square elements for deep localized hyperthermia. These array include a 20x20 planar array, a 80x16 cylindrical-section array, and a 16x16 spherical-section array. Also, we have designed a new phased array with circular elements. We used single and multiple focus scanning methods with intensity gain maximization. We found that the array with circular elements is more effective in reducing grating lobes compared to the same array with square elements. The grating lobes were at least 30 dB smaller than the phased array with square elements and intensity gain was at least 1 dB greater. In addition, with equal intensity distribution patterns for rectangular and circular phased arrays, the number of elements in the circular phased array was smaller and its intensity gain was greater than the other arrays.

The effect of lossy, MP3 (MPEG-Layer 3) compression on clinically important Doppler parameters - derived from spectral analysis of Doppler ultrasound signals - was investigated. Ten, 10-second acquisitions of gated Doppler ultrasound signal were collected in a phantom perfused with a pulsatile flow waveform. Doppler data were collected using two sample volume lengths - 1.5 mm and 10 mm. The in- phase and quadrature Doppler signals were digitized at 44.1 kHz and compressed using four grades of signal compression (with corresponding compression ratios given in brackets): uncompressed, 128 kbits/s (11:1), 64 kbits/s (44:1). The digital audio signals were identically processed with a Fourier analysis program that provided an estimate of the instantaneous Doppler frequency (velocity) spectrum and derived parameters such as peak velocity, mean velocity, spectral width, total integrated power, and ratio of spectral power from negative and positive velocities. Analysis of variance indicated there were no significant differences (p>0.05) observed in the peak or mean velocities, spectral width, or the power ratio derived from 128 kbits/s and 64 kbits/s audio signals when compared to the uncompressed audio signals (both sample volume lengths) and the 128 kbits/s audio signals (10 mm sample volume length). However, for the 32 kbits/s audio signals, significant differences (p<0.001) were found in all of the studied parameters.

This work aims at developing innovative algorithms and tools for summarizing echocardiogram videos. Specifically, we summarize the digital echocardiogram videos by temporally segmenting them into the constituent views and representing each view by the most informative frame. For the segmentation we take advantage of the well-defined spatio- temporal structure of the echocardiogram videos. Two different criteria are used: presence/absence of color and the shape of the region of interest (ROI) in each frame of the video. The change in the ROI is due to different modes of echocardiograms present in one study. The representative frame is defined to be the frame corresponding to the end- diastole of the heart cycle. To locate the end-diastole we track the ECG of each frame to find the exact time the time- marker on the ECG crosses the peak of the end-diastole we track the ECG of each frame to find the exact time the time- marker on the ECG crosses the peak of the R-wave. The corresponding frame is chosen to be the key-frame. The entire echocardiogram video can be summarized into either a static summary, which is a storyboard type of summary and a dynamic summary, which is a concatenation of the selected segments of the echocardiogram video. To the best of our knowledge, this if the first automated system for summarizing the echocardiogram videos base don visual content.

An approach for acquiring dimensionally accurate 3D ultrasound data, based on a modified 1D transducer array, is presented. Th method avoids many of the drawbacks of conventional approaches to 3D ultrasound data acquisition. Scanning is simple and easy to perform in a clinical setting. A modified 1D transducer array is employed comprising a central conventional 1D imaging array and two perpendicular tracking arrays - each integrally mounted at each end of the imaging array. As the transducer is scanned in the elevation direction of the central array, the images acquired by the tracking array remain coplanar and hence it is possible to accurately track image motion using any one of several image tracking techniques. Methods for improving the performance and ergonomics of the transducer array are presented. In particular, a crossed electrode transducer structure is proposed for minimizing the total transducer footprint (contact surface area). The versatility of the approach in terms of its suitability for scanning breast, carotid and prostate is discussed. We have acquired both phantom and in-vivo 3D ultrasound data with the prototype imaging approach. Initial studies suggest that the linear dimensional accuracy in the elevation direction (i.e., the reconstructed direction) is approximately 5%.

Understanding hemodynamics is essential for understanding vascular disease development, accurately diagnosing the disease, and studying disease progression. A vector Doppler system has been developed and used to conduct detailed studies of the post stenotic flow field in hopes of furthering this understanding. This paper presents the initial experiments on a steady flow model, specifically, the results for a 40% diameter reducing stenosis with flow rates ranging from 10 cm3/s to 75 cm3/s. The centerstream jet, recirculation zone, shear layer, and turbulent jet break down could all be delineated. The variance in flow velocity, evaluated through spectral broadening bandwidth and turbulent intensity, provided an alternative approach to imaging the flow downstream of a stenosis. The downstream position of maximum turbulent intensity was also identified, and an analysis of eddies present within the shear layer conducted. Each of these parameters potentially contributes toward identification of the initiation, extent, and progression of plaque formation, as well as the interrelation between hemodynamics and vascular remodeling.

Assessment of intramyocardial perfusion by contrast echocardiography is a promising new technique that allows to obtain quantitative parameters for the assessment of ischemic disease. In this work, a new methodology and a software prototype developed for this task are presented. It has been validated with Coherent Contrast Imaging (CCI) images acquired with an Acuson Sequoia scanner. Contrast (Optison microbubbles) is injected continuously during the scan. 150 images are acquired using low mechanical index U/S pulses. A burst of high mechanical index pulses is used to destroy bubbles, thus allowing to detect the contrast wash-in. The stud is performed in two conditions: rest and pharmacologically induced stress. The software developed allows to visualized the study (cine) and to select several ROIs within the heart wall. The position of these ROIs along the cardiac cycle is automatically corrected on the basis of the gradient field, and they can also be manually corrected in case the automatic procedure fails. Time curves are analyzed according to a parametric model that incorporates both contrast inflow rate and cyclic variations. Preliminary clinical results on 80 patients have allowed us to identify normal and pathological patterns and to establish the correlation of quantitative parameters with the real diagnosis.

The purpose of this system is to have the capability to characterize the performance of very high frequency transducers and arrays. The analog front end is computer controlled by a set of de-multiplexers and multiplexers. The output of the multiplexer network is connected to a TGC array, which is interfaced to a high-speed data acquisition system. A software GUI (Graphical User Interface) has been design to accomplish this task. A programmable digital I/O interface allows collection of RF channel data and has the capability to be interfaced to a very high frequency analog beamformer under construction. The system front-end electronics (pulsers, receivers, T/R switches, multiplexers, and demultiplexers) have been characterized. The digital I/O signal interface has been integrated and tested. The hardware front end has been integrated to the array interface distribution panel. The individual transducer elements impulse responses have been evaluated and the performance of the array has been tested with a wire test phantom to characterize lateral and axial resolution.

We are developing an acoustic levitation chamber for measuring adhesion force strengths among biological cells. Our research has four phases. Phase I, presented here, is concerned with the design and construction of a chamber for trapping cell-sized microbubbles with known properties in acoustic standing waves, and examines the theory that describes the standing wave field. A cylindrical chamber has been developed to generate a stable acoustic standing wave field. The pressure field was mapped using a 0.4-mm needle hydrophone, and experiments were performed using 100 micron diameter unencapsulated air bubbles, 9 micron diameter isobutane-filled microbubbles, and 3 micron diameter decafluorobutane (C₄F₁₀)-filled microbubbles, confirming that the net radiation force from the standing wave pressure field tends to band the microbubbles at pressure antinodes, in accordance with theory.

Strain Rate (SR) Imaging is a recent imaging technique that provides information about regional myocardial deformation by measuring local compression and expansion rates. SR can be obtained by calculating the local in-plane velocity gradients along the ultrasound beam from Doppler Tissue velocity data. However, SR calculations are very dependent on the image noise and artifacts, and different calculation algorithms may provide inconsistent results. This paper compares techniques to calculate SR. 2D Doppler Tissue Images (DTI) are acquired with an Acuson Sequoia scanner. Noise was measured with the aid of a rotating phantom. Processing is performed on polar coordinates. For each image, after removal of black spot artifacts by a selective median filter, two different SR calculation methods have been implemented. In the first one, SR is computed as the discrete velocity derivative, and noise is reduced with a variable-width gaussian filter. In the second method a smoothing cubic spine is calculated for every scan line according to the noise level and the derivative is obtained from an analytical expression. Both methods have been tested with DTI data from synthetic phantoms and normal volunteers. Results show that noise characteristics, border effects and the adequate scale are critical to obtain meaningful results.

One of the main problems in the study of complex systems is to define a good metric that can distinguish between different dynamical behaviors in a nonlinear system. In this work we describe a method to detect different types of behaviors in a long term ECG-Holter using short portions of the Holter signal. This method is based on the calculation of the statistical distance between two distributions in a phase-space of a dynamical system. A short portion of an ECG-Holter signal with normal behavior is used to reconstruct the trajectory of an attractor in low dimensional phase-space. The points in this trajectory are interpreted as statistical distributions in the phase-space and assumed to represent the normal dynamical behavior of the ECG recording in this space. A fast algorithm is then used to compute the statistical distance between this attractor and all other attractors that are built using a sliding temporal window over the signal. For normal cases the distance stayed almost constant and below a threshold. For cases with abnormal transients, on the abnormal portion of ECG, the distance increased consistently with morphological changes.

Microwave thermotherapy (MT) is an oncological treatment. At present the invasive thermometer probes are clinically used for temperature measuring during an MT. Any invasive handling of tumors is of high-risk. A new method of noninvasive monitoring of temperature distribution in tissue has been developed. An MT treatment of the experimentally induced pedicle-tumors of the rat was prepared. It consists of an intelligent regulation loop controlling a high frequency (HF) generator according to the maximal measured temperature in the tissue and a special HF MT applicator. The loop is also equipped with an invasive thermometer (4 invasive probes). During the MT treatment the series of ultrasound B-mode images were obtained. The texture parameters were evaluated form the obtained ultrasound images. These parameters were correlated with the invasively measured temperature during the MT session. For 60 rat samples a strong correlation between the mean gray level in the ROIs in the ultrasound pictures and the invasively measured temperature in the range 37-44 degree(s)C (98.6-111.2 F) was found. The correlation coefficient of the mean gray level and the invasively measured temperature is 0.96+/- 0.05. A system for representation of changes of spatial temperature distribution of the whole tumor during MT will be presented.