Ultrasonic imaging began, like life, in the sea, with the development of sonar for detecting submarines after World-War 1. However, to begin to image soft tissues the ranging time of ocean sonars needed to be reduced, and the electronics speeded up, by a factor of about the ratio between nautical miles and centimeters. This was only possible after the electronic developments made for radar in World-War 2. The rest of our technical history closely follows the developments in semiconductors and fabrication methods that led to modern electronics. This is a largely personal story of a recently graduated engineer with radar experience, who began with fabricating equipment to be used in the hospital to diagnose breast cancer, and continued with involvement the development of echocardiography and Doppler devices. Along the way many others have contributed to the field, including work in other countries that is not covered here. In future, ultrasonic imaging may hold the key to understanding some fundamental questions in human health if adopted for screening studies. It alone offers a relatively inexpensive imaging method that is free of known hazards.

A method is described for repeatably assessing elasticity and 3D extent of suspected prostate cancers. Elasticity is measured by controlled water inflation of a sheath placed over a modified transrectal ultrasound transducer. The benefit of using fluid inflation is that it should be possible to make repeatable, accurate, measurements of elasticity that are of interest in the serial assessment of prostate cancer progression or remission. The second aspect of the work uses auxiliary tracking arrays placed at each end of the central imaging array that allow the transducer to be rotated while simultaneously collected 'tracking' information thus allowing the position of successive image planes to be located with approximately 11% volumetric accuracy in 3D space. In this way, we present a technique for quantifying volumetric extent of suspected cancer in addition to making measures of elastic anomalies.

Purpose: Harmonic imaging has become a well-established technique for ultrasonic imaging at fundamental frequencies of 10 MHz or less. Ophthalmology has benefited from the use of fundamentals of 20 MHz to 50 MHz. Our aim was to explore the ability to generate harmonics for this frequency range, and to generate harmonic images of the eye. Methods: The presence of harmonics was determined in both water and bovine vitreous propagation media by pulse/echo and hydrophone at a series of increasing excitation pulse intensities and frequencies. Hydrophone measurements were made at the focal point and in the near- and far-fields of 20 MHz and 40 MHz transducers. Harmonic images of the anterior segment of the rabbit eye were obtained by a combination of analog filtering and digital post-processing. Results: Harmonics were generated nearly identically in both water and vitreous. Hydrophone measurements showed the maximum second harmonic to be -5 dB relative to the 35 MHz fundamental at the focus, while in pulse/echo the maximum harmonic amplitude was -15dB relative to the fundamental. Harmonics were absent in the near-field, but present in the far-field. Harmonic images of the eye showed improved resolution. Conclusion: Harmonics can be readily generated at very high frequencies, and at power levels compliant with FDA guidelines for ophthalmology. This technique may yield further improvements to the already impressive resolutions obtainable in this frequency range. Improved imaging of the macular region, in particular, may provide significant improvements in diagnosis of retinal disease.

Progress made in the development of a miniature real-time volumetric ultrasound imaging system is presented. This system is targeted for use in a 5-mm endoscopic channel and will provide real-time, 30-mm deep, volumetric images. It is being developed as a clinically useful device, to demonstrate a means of integrating the front-end electronics with the transducer array, and to demonstrate the advantages of the capacitive micromachined ultrasonic transducer (CMUT) technology for medical imaging. Presented here is the progress made towards the initial implementation of this system, which is based on a two-dimensional, 16x16 CMUT array. Each CMUT element is 250 um by 250 um and has a 5 MHz center frequency. The elements are connected to bond pads on the back side of the array with 400-um long through-wafer interconnects. The transducer array is flip-chip bonded to a custom-designed integrated circuit that comprises the front-end electronics. The result is that each transducer element is connected to a dedicated pulser and low-noise preamplifier. The pulser generates 25-V, 100-ns wide, unipolar pulses. The preamplifier has an approximate transimpedance gain of 500 kOhm and 3-dB bandwidth of 10 MHz. In the first implementation of the system, one element at a time can be selected for transmit and receive and thus synthetic aperture images can be generated. In future implementations, 16 channels will be active at a given time. These channels will connect to an FPGA-based data acquisition system for real-time image reconstruction.

We have designed, fabricated, and characterized two-dimensional 16x16-element capacitive micromachined ultrasonic transducer (CMUT) arrays. The CMUT array elements have a 250-μm pitch, and when tested in immersion, have a 5 MHz center frequency and 99% fractional bandwidth. The fabrication process is based on standard silicon micromachining techniques and therefore has the advantages of high yield, low cost, and ease of integration. The transducers have a Si3N4 membrane and are fabricated on a 400-μm thick silicon substrate. A low parasitic capacitance through-wafer via connects each CMUT element to a flip-chip bond pad on the back side of the wafer. Each through wafer via is 20 μm in diameter and 400 μm deep. The interconnects form metal-insulator-semiconductor (MIS) junctions with the surrounding high-resistivity silicon substrate to establish isolation and to reduce parasitic capacitance. Each through-wafer via has less than 0.06 pF of parasitic capacitance. We have investigated a Au-In flip-chip bonding process to connect the 2D CMUT array to a custom integrated circuit (IC) with transmit and receive electronics. To develop this process, we fabricated fanout structures on silicon, and flip-chip bonded these test dies to a flat surface coated with gold. The average series resistance per bump is about 3 Ohms, and 100% yield is obtained for a total of 30 bumps.

A kerfless eight element high frequency ultrasound annular array transducer using 9 μm P(VDF-TrFE) bonded to a high density flexible interconnect was fabricated. The flexible circuit composed of Kapton polyimide film with gold electrode pattern of equal area annuli apertures on the top side of a 50 μm thick Kapton polyimide film. Each element had several 30 μm diameter electroplated vias that connected to electrode traces on the bottom side of the Kapton polyimide film. There was a 30 μm spacing between elements. The total aperture of the array was 3.12 mm. The transducer's performance has been modeled by implementing the Redwood version of the Mason model into PSpice and using the Krimholtz, Leedom and Matthaei (KLM) model utilized in the commercial software PiezoCAD. The transducer’s performance was evaluated by measuring the electrical impedance with a HP 4194 impedance analyzer, pulse echo response using a Panametrics 5900 pulser/receiver and crosstalk measurement for each element in the array. The measured electrical impedance for each element was 540 Ω and -84° phase. In order to improve device sensitivity an inductor was attached in series with each element to reduce the insertion loss to 33 dB. The measured average center frequency and bandwidth of each element was 55 MHz and 50% respectively. The measured crosstalk at the center frequency was -45 dB in water.

Capacitive micromachined ultrasonic transducer (cMUT) technology has been recognized as an attractive alternative to the more traditional piezoelectric transducer technology in medical ultrasound imaging for several years now. There are mainly two reasons for the interest in this technology: Micromachining is derived from the integrated circuit technology and therefore shares the well-known advantages and experience of it. Also, capacitive transduction using thin membranes has fundamental superiorities over the piezoelectric transduction mechanism such as wide frequency bandwidth. Capacitive micromachined ultrasonic transducers are essentially capacitor cells where the two plates of the capacitor, the membrane and the substrate, are separated with a vacuum sealed cavity. Typically, a cMUT is made of many micro-scale capacitor cells operating in parallel. This paper describes a new fabrication technique for building cMUTs which is called the wafer-bonding method. In this method, the cavity and the membrane are defined on separate wafers and brought together by wafer-bonding in vacuum. The wafer-bonding method has several advantages over the traditional sacrificial release method of cMUT fabrication. It allows greater flexibility in the cMUT design which means better device performance. It reduces the number of process steps, device turn-around time, and increases the overall uniformity, reliability. and repeatability. Device examples of one-dimensional and two-dimensional arrays designed to work in the 1 to 50 MHz range with 100% fractional bandwidth highlight the advantages of this method, and show that cMUT technology is indeed the better candidate for next generation ultrasonic imaging arrays.

The purpose of this study is to investigate the imaging capability of a CMOS (PE-CMOS) ultrasound sensing array coated with piezoelectric material. There are three main components in the laboratory setup: (1) a transducer operated at 3.5MHz-7MHz frequency generating unfocused ultrasound plane waves, (2) an acoustic compound lens that collects the energy and focuses ultrasound signals onto the detector array, and (3) a PE-CMOS ultrasound sensing array (Model I400, Imperium Inc. Silver Spring, MD) that receives the ultrasound and converts the energy to analog voltage followed by a digital conversion. The PE-CMOS array consists of 128×128 pixel elements with 85μm per pixel. The major improvement of the new ultrasound sensor array has been in its dynamic range. We found that the current PE-CMOS ultrasound sensor (Model I400) possesses a dynamic range up to 70dB. The system can generate ultrasound attenuation images of soft tissues which are similar to digital images obtained from an x-ray projection system. In the paper, we also show that the prototype system can image bone fractures using reflective geometry.

Transducers based on piezoelectric crystals dominate the biomedical ultrasonic imaging field. However, fabrication difficulties for piezoelectric transducers limit their usage for complex imaging modalities such as 2D imaging, high frequency imaging, and forward looking intravascular imaging. Capacitive micromachined ultrasonic transducers (CMUTs) have been proposed to overcome these limitations and they offer competitive advantages in terms of bandwidth and dynamic range. Further, the ease of fabrication enables manufacturing of complex array geometries. A CMUT transducer is composed of many electrostatically actuated membranes. Earlier analysis of these devices concentrated on an equivalent circuit approach, which assumed the motion of the membrane was approximated by a parallel plate capacitor. Finite element analysis is required for more accurate results. In this paper, we present the finite element model developed to evaluate the performance of the CMUTs. The model is composed of a membrane radiating into immersion medium. Electrostatic actuation is added on using electromechanical elements. Symmetry boundary conditions are imposed around the sidewalls of the finite element mesh, so that the model reflects the properties of a cell driven with the same phase as its neighboring membranes in an infinitely large array. Absorbing boundaries are implemented one wavelength away from the membrane to avoid reflections from the end of the finite element mesh. Using the model, we optimized the membrane radius, membrane thickness and gap height. Our optimized designed yielded a center frequency of 13 MHz with hundred percent bandwidth. A maximum output pressure of 20 kPascal per volt was obtained.

Applications of ultrasonic imaging in fields such as dermatology, ophthalmology, and cardiovascular medicine require very high resolution. Limitations in existing transducer technologies inhibit the development of high-frequency arrays, which would allow the use of dynamic focusing and enable higher frame rates. As an alternative, capacitive micromachined ultrasonic transducer (CMUT) technology, using integrated circuit fabrication techniques, can provide arrays with the small dimensions required for high-frequency operation. We have designed and fabricated several linear and ring arrays of CMUTs to operate in the 10 to 50 MHz range. These new arrays are made with the wafer bonding process. The ring arrays in particular demonstrate the feasibility of thinning the transducer to aid packaging in intravascular applications. This study shows that CMUTs can be made for high-frequency operation. Both transducers for use in conventional and collapse-mode operation have been designed and characterized. The results demonstrate that CMUT is an appropriate technology for building high-frequency arrays. A linear array of high-voltage pulser and amplifier circuits has also been designed for use with an array of CMUTs to enable real-time imaging applications. Pulse-echo results from the sixteen-channel array have been demonstrated.

Power Doppler ultrasound enhanced by microbubble contrast agent has been used to image tissue vascularity and blood flow for the assessment of antivascular therapies. We have proposed a multigating technique that measures bubble concentration as a function of ultrasound exposure for deriving tumor blood flow and vascularity.1 Techniques using ultrasound contrast agent are known to be sensitive to the choice of imaging parameters like mechanical index and tissue attenuation. In this paper, the roles of mechanical index (MI) and tissue attenuation were evaluated experimentally in a rubber tubing flow phantom connected to a mixing chamber and a variable speed pump. The contrast was injected in the mixing chamber and the flow rate was measured using power Doppler imaging. The measurements were repeated at different MIs (0.1 to 1.3), and at different levels of attenuation, obtained with solutions of glycerol-water (10-20%). True flow was measured by collecting liquid flowing out of the phantom over a fixed duration. At low MI (<0.5), the grayscale and Doppler signal were weak, making these images unsuitable for analysis. At higher MI (> 0.8), there was a well-defined enhancement by contrast agent resulting in reproducible flow measurements at variable MIs. A balance between the number of bubbles destroyed and the echo they generate must be achieved for optimal imaging. The increased attenuation of ultrasound by the overlying medium did not influence the flow measurements.

An intact mouse model of surgically-induced myocardial infarction (MI) caused by permanent occlusion of the Left Anterior Descending (LAD) coronary artery was studied. Normal mice with no occlusion were also studied as controls. For each mouse, contrast enhanced ultrasound images of the heart were acquired in parallel cross-sections perpendicular to the sternum at millimeter increments. For accurate 3D reconstruction, ECG gating and a tri-axial adjustable micromanipulator were used for temporal and spatial registration. Ultrasound images at steady-state of blood refilling were color-coded in each slice to show relative perfusion. Myocardial perfusion defects and necrosis were also examined postmortem by staining with Phthalo blue and TTC red dyes. Good correlation (R>0.93) in perfused area size was observed between in vivo measurements and histological staining. A 3D multi-slice model and a 3D rendering of perfusion distribution were created and showed a promising match with postmortem results, lending further credence to its use as a more comprehensive and more reliable tool for in vivo assessment of myocardial perfusion than 2D tomographic analysis.

It is important to understand the interaction of pulsed ultrasound with ultrasound contrast agents so that the agents can be utilized to their fullest. Towards this goal, we have optimized light-scattering to provide quantitative information about microbubble oscillations when subjected to diagnostic ultrasound pulses. Initial experiments were performed with individual microbubbles. Pulses from actual diagnostic imaging systems were used to 'activate' the microbubbles. Bubble oscillations were measured by focusing the scattered light onto a photodetector. Data was collected with a fast oscilloscope set up to capture instantaneous bubble oscillation data for many consecutive ultrasound pulses. Results for individual bubbles show that at low pressures, the bubbles can be stable for many pulses. Sometimes, the bubbles evolve, presumably due to shell permeabilization and shell fatigue. Bubble dynamics models compare favorably with the observed oscillations. Light scattering can be an important tool for understanding and optimizing newer bubble imaging modes such as pulse inversion.

This paper describes our "virtual" Doppler ultrasound (DUS) system, in which colour DUS (CDUS) images and DUS spectrograms are generated on-the-fly and displayed in real-time in response to position and orientation cues provided by a magnetically tracked handheld probe. As the presence of complex flow often confounds the interpretation of Doppler ultrasound data, this system will serve to be a fundamental tool for training sonographers and gaining insight into the relationship between ambiguous DUS images and complex blood flow dynamics. Recently, we demonstrated that DUS spectra could be realistically simulated in real-time, by coupling a semi-empirical model of the DUS physics to a 3-D computational fluid dynamics (CFD) model of a clinically relevant flow field. Our system is an evolution of this approach where a motion-tracking device is used to continuously update the origin and orientation of a slice passing through a CFD model of a stenosed carotid bifurcation. After calibrating our CFD model onto a physical representation of a human neck, virtual CDUS images from an instantaneous slice are then displayed at a rate of approximately 15 Hz by simulating, on-the-fly, an array of DUS spectra and colour coding the resulting spectral mean velocity using a traditional Doppler colour scale. Mimicking a clinical examination, the operator can freeze the CDUS image on-screen, and a spectrogram corresponding to the selected sample volume location is rendered at a higher frame rate of at least 30 Hz. All this is achieved using an inexpensive desktop workstation and commodity graphics card.

We have developed a new adaptive clutter rejection technique where an optimum clutter filter is dynamically selected according to the varying clutter characteristics in ultrasound color Doppler imaging. The selection criteria have been established based on the underlying clutter characteristics (i.e., the maximum instantaneous clutter velocity and the clutter power) and the properties of various candidate clutter filters (e.g., projection-initialized infinite impulse response and polynomial regression). We obtained an average improvement of 3.97 dB and 3.27 dB in flow signal-to-clutter-ratio (SCR) compared to the conventional and down-mixing methods, respectively. These preliminary results indicate that the proposed adaptive clutter rejection method could improve the sensitivity and accuracy in flow velocity estimation for ultrasound color Doppler imaging. For a 192 x 256 color Doppler image with an ensemble size of 10, the proposed method takes only 57.2 ms, which is less than the acquisition time. Thus, the proposed method could be implemented in modern ultrasound systems, while providing improved clutter rejection and more accurate velocity estimation in real time.

The harmonic behavior of a ultrasound contrast agent (2nd generation) begins at a peak negative pressure of 10 kPa and finishes approximately at 60 kPa with the rupture of the contrast agent's bubble. Moreover, increasing the power of the ultrasound pulse, the tissue begins to have a har-monic response too, affecting the imaging of the contrast agent. The survival of the bubbles is affected by different parameters; the most important is the intensity of the ultrasound beam which can be related to several index: peak negative pressure, acoustic intensity (Ispta) and the mechanical index (MI). Therefore with harmonic imaging it is important to use low power pulses with a good accuracy and reproducibility; in order to optimize this technique is necessary to find a good index of the probability of destruction of the bubbles. Most of the scanners use the display of the mechanical index as the parameter associated to the harmonic behavior of the contrast agent. A special phantom was realized in order to measure the MI of different probes of different scanners with a hydrophone. Aim of this study is to evaluate the accuracy of the MI displayed by different scanners and to verify its correlation with the others parameter related to the microbubbles persistence.

The goal of this study is to evaluate an Artificial Neural Network (ANN) for differentiating benign and malignant breast masses on ultrasound scans. The ANN was designed with three layers (input, hidden and output layer), where a sigmoidal (hyperbolic tangent) response function is used as an activation function at each unit. Data from 54 patients with biopsy-proven malignant (N=20) and benign (N=34) masses were used to evaluate the diagnostic performance of the ANN. Of the seven quantitative features extracted from ultrasound images, only four showed statistically significant difference between the two categories. These features were margin sharpness, margin echogenicity, angular continuity, and age of patients. The diagnostic performance was evaluated by round-robin substitution to negate bias due to small sample size. All the input features were standardized to zero-mean and unit-variance to prevent non-uniform learning, which can generate unwanted error. The outputs of the network were analyzed by Receiver Operating Characteristics (ROC). The resulting area under the ROC curve Az was 0.856 ± 0.058 with 95% confidence limit from 0.734 to 0.936, providing 76.5% specificity at 95% sensitivity. The performance of the ANN was comparable to the performance by logistic regression analysis reported by our group earlier. These results suggest that an ANN when combined with sonography can effectively classify malignant and benign breast lesions.

Endothelial dysfunction in response to vasoactive stimuli is closely associated with diseases such as atherosclerosis, hypertension and congestive heart failure. The current method of using ultrasound to image the brachial artery along the longitudinal axis is insensitive for measuring the small vasodilatation that occurs in response to flow mediation. The goal of this study is to overcome this limitation by using cross-sectional imaging of the brachial artery in conjunction with the User-Guided Automated Boundary Detection (UGABD) algorithm for extracting arterial boundaries. High-resolution ultrasound imaging was performed on rigid plastic tubing, on elastic rubber tubing phantoms with steady and pulsatile flow, and on the brachial artery of a healthy volunteer undergoing reactive hyperemia. The area of cross section of time-series images was analyzed by UGABD by propagating the boundary from one frame to the next. The UGABD results were compared by linear correlation with those obtained by manual tracing. UGABD measured the cross-sectional area of the phantom tubing to within 5% of the true area. The algorithm correctly detected pulsatile vasomotion in phantoms and in the brachial artery. A comparison of area measurements made using UGABD with those made by manual tracings yielded a correlation of 0.9 and 0.8 for phantoms and arteries, respectively. The peak vasodilatation due to reactive hyperemia was two orders of magnitude greater in pixel count than that measured by longitudinal imaging. Cross-sectional imaging is more sensitive than longitudinal imaging for measuring flow-mediated dilatation of brachial artery, and thus may be more suitable for evaluating endothelial dysfunction.

With relatively high frame rates and the ability to acquire volume data sets with a stationary transducer, 3D ultrasound systems, based on matrix phased array transducers, provide valuable three-dimensional information, from which quantitative measures of cardiac function can be extracted. Such analyses require segmentation and visual tracking of the left ventricular endocardial border. Due to the large size of the volumetric data sets, manual tracing of the endocardial border is tedious and impractical for clinical applications. Therefore the development of automatic methods for tracking three-dimensional endocardial motion is essential. In this study, we evaluate a four-dimensional optical flow motion tracking algorithm to determine its capability to follow the endocardial border in three dimensional ultrasound data through time. The four-dimensional optical flow method was implemented using three-dimensional correlation. We tested the algorithm on an experimental open-chest dog data set and a clinical data set acquired with a Philips' iE33 three-dimensional ultrasound machine. Initialized with left ventricular endocardial data points obtained from manual tracing at end-diastole, the algorithm automatically tracked these points frame by frame through the whole cardiac cycle. A finite element surface was fitted through the data points obtained by both optical flow tracking and manual tracing by an experienced observer for quantitative comparison of the results. Parameterization of the finite element surfaces was performed and maps displaying relative differences between the manual and semi-automatic methods were compared. The results showed good consistency between manual tracing and optical flow estimation on 73% of the entire surface with fewer than 10% difference. In addition, the optical flow motion tracking algorithm greatly reduced processing time (about 94% reduction compared to human involvement per cardiac cycle) for analyzing cardiac function in three-dimensional ultrasound data sets.

Minimally invasive surgery in combination with ultrasound (US) imaging imposes high demands on the surgeon's hand-eye-coordination capabilities. A possible solution to reduce these requirements is minimally invasive robotic surgery in which the instrument is guided by visual servoing towards the goal defined by the surgeon in the US image. This approach requires robust tracking of the instrument in the US image sequences which is known to be difficult due to poor image quality. This paper presents algorithms and results of first tracking experiments. Adaptive thresholding based on Otsu's method allows to cope with large intensity variations of the instrument echo. Median filtering of the binary image and subsequently applied morphological operations suppress noise and echo artefacts. A fast run length code based labelling algorithm allows for real-time labelling of the regions. A heuristic exploiting region size and region velocity helps to overcome ambiguities. The overall computation time is less than 20 ms per frame on a standard PC. The tracking algorithm requires no information about texture and shape which are known to be very unreliable in US image sequences. Experimental results for two different instrument materials (polyvinyl chloride and polyurethane) are given, showing the performance of the proposed approach. Choosing the appropriate material, trajectories are smooth and only few outliers occur.

Computer-aided characterization of a breast ultrasound lesion involves two steps: first, extracting features from the lesion whose boundary is pre-defined on the images, and then converting the features into mathematical models. Most methods assume that the boundaries of the lesions are pre-selected or outlined by sonographers or operators, because automated delineation of lesion boundary is not trivial and is a challenging task. The purpose of this study was to develop and evaluate an automated lesion boundary segmentation method that is based on texture-based, multi-resolution image analysis. One hundred ninety-seven breast ultrasound images containing solid breast lesions from 172 women (age 24-89 years, mean 38 years) were studied. Fifty-five of the 197 images were from 40 women with malignant lesions, and the remaining 142 were from 132 patients with benign lesions. Each breast lesion was identified by an operator who placed a rectangular region of interest (ROI) to widely encompass the lesion. The resolution of the image was compressed, at variable ratios depending on the ROI size, to reduce noise. Texture momentum was computed. A binary image was generated from the texture and pixel intensity parameters. Initial seed boundary was segmented from the binary image and then expanded to the original resolution using the boundary pixel intensity gradient information. The boundary of each breast lesion was delineated by a breast-imaging radiologist who was blinded to the computer-detected lesion boundary. The 'area match ratio' between the manually drawn boundaries and the automatically detected boundaries was computed. This ratio is equal to or less than unity (unity indicates that the areas match exactly). Overall, good agreement was seen between the multi-resolution segmentation method and the radiologist’s manual delineation. The mean area match ratio was 0.87 ±0.02. We have developed a multi-resolution, texture-based method to segment the boundary of breast lesions. This method will facilitate full automation for the characterization of breast ultrasound lesions.

We propose a way of measuring elastic properties of tissues in-vivo, using standard medical image ultrasound machine without any special hardware. Images are acquired while the tissue is being deformed by a varying pressure applied by the operator on the hand-held ultrasound probe. The local elastic shear modulus is either estimated from a local displacement field reconstructed by an elastic registration algorithm, or both the modulus and the displacement are estimated simultaneously. The relation between modulus and displacement is calculated using a finite element method (FEM). The estimation algorithms were tested on both synthetic, phantom and real subject data.

Acoustic radiation force impulse (ARFI) imaging has been demonstrated to provide insight into the mechanical properties of tissue. The quality of ARFI images is dependent on the amount of acoustic energy from the radiation force pulse reaching the focus. Intra-cardiac probes provide an advantage for ARFI imaging of cardiac tissue, as the probe can be positioned close to the region of interest. The resulting ARFI images display local variations in tissue stiffnesses and show promise for monitoring and assessing the progress of cardiac ablations. The Siemens AcuNav intra-cardiac probe was used to image a tissue-mimicking phantom having 3 mm diameter spherical inclusions with an elastic modulus eight times greater than the surrounding tissue. The ARFI sequences formed high contrast, high resolution images of these inclusions up to depths of approximately 1.5 cm. The ARFI pulse sequences resulted in 0.8°C temperature increase on the transducer face, and the time constant associated with the return to equilibrium temperature was approximately 300 ms. The probe was used to examine an excised segment of an ovine right ventricle with a surface lesion created from radiofrequency ablations (RFA). In areas of healthy tissue, the ARFI images did not show any stiffer regions that would indicate the presence of a lesion. Although the lesion was not visible in conventional B-mode images, the ARFI images were able to show the boundaries between the lesion and the surrounding tissue.

Acoustic Radiation Force Impulse (ARFI) imaging utilizes brief, high-energy acoustic pulses to excite tissue and ultrasonic correlation based tracking methods to monitor the resulting tissue displacement, which reflects the relative mechanical properties of tissue (i.e. stiffer tissue displaces less). ARFI image contrast is optimized utilizing tightly focused radiation force excitations at multiple axial and lateral locations throughout a 2D field of view. In an ongoing, IRB approved, clinical study, suspicious breast lesions are interrogated in vivo via multi-focal-zone ARFI prior to undergoing core biopsy. A Siemens SONOLINE Antares (TM) scanner and VF10-5 probe were configured to acquire ARFI data from multiple focal-zones and lateral locations. Data was acquired in real-time, and processed off-line. Processing included: filtering, parametric data analysis, normalization and combination of the multiple focal-zone data, and automatic edge detection. ARFI sequences were designed with varying pushing pulse frequencies and intensities. Contrast to noise ratio was evaluated in a tissue mimicking phantom for lesions at different depths using the different pushing pulse sequences. For shallower lesions (depth=10mm), CNR was higher than for deeper lesions, and did not vary appreciably for the different push sequences. For deeper lesions (depth=20mm), CNR increased with increasing push pulse intensity and decreasing push pulse frequency. With the pushing pulse transmit intensity calibrated (in a homogeneous phantom) to achieve uniform displacement at all axial depths, in vivo results yielded poor SNR at depth and did not achieve overall uniform displacement. In vivo, image quality improved with increasing push pulse intensity. To date, 27 masses have been interrogated using multi-focal-zone ARFI and overall good structural agreement exists between B-mode and ARFI images. Normalization and blending facilitate image generation from ARFI interrogation using different intensities at different focal depths.

An elasticity imaging system was developed for measuring the stiffness variation at different depths of the human fingerpad skin in vivo. In this system, ultrasonic backscatter microscopy (UBM) with a single high frequency (28MHz) transducer was employed to obtain data on tissue heterogeneity at high axial resolution (~25 mm). The dorsal side of the finger was fixed on a manually controlled vertical stage and an acrylic indentor was applied to the fingerpad. A slit cut vertically through the indentor at the center and a piece of transparency sheet attached to the bottom allowed most of the ultrasound power to pass though while maintaining a flat surface in contact with the skin. With the assumption that the skin can be modeled as a semi-infinite layered structure, only data from a single A-line was obtained for strain analysis. The data at continuous indentation steps were cross-correlated to calculate the displacement at different spots along the depth. The de-correlation at certain regions was resolved by removing the data points with lower correlation coefficients, and curve fitting was applied to overcome the lack of resolution due to sampling. The fingerpads of 10 human subjects were tested in vivo and a gelatin phantom was made and tested for comparison. The results showed that even though some data were degraded due to the hypoechoic nature of the subcutaneous fat, the axial strain profile through the skin thickness (up to 3mm in depth) could be extracted as a measure of the stiffness variation.

Acoustic Radiation Force Impulse (ARFI) imaging uses short duration acoustic pulses to generate and subsequently determine localized displacements in tissue. Time delay estimators, such as normalized cross correlation and phase shift estimation, form the computational basis for ARFI imaging. This paper considers these algorithms and the effects of noise, interpolation, and quadrature demodulation on the accuracy of the time delay estimates. These results are used to implement a real-time ARFI imaging system and in an ex vivo liver ablation study.

Increased stiffness of the arteries has recently gained acceptance as an independent risk factor for cardiovascular and many other diseases. Pulse wave velocity (PWV) is widely used for estimating the stiffness of an artery. From measured PWV, the diameter and thickness of the artery are needed to calculate the elastic modulus of the artery. In this paper a new method of using ring resonant mode for estimation of arterial elastic modulus is proposed. To generate the ring resonance, a localized radiation force of ultrasound is remotely and non-invasively applied at the artery. The vibration response of the artery is measured by optical or Doppler techniques. Three ring resonant modes are identified for estimation of the elastic modulus of the artery. Experiments on ring resonant frequencies were carried out on a porcine artery. The measurement results of the three resonant frequencies are, respectively, 356 Hz, 718 Hz, and 968 Hz. Estimation of the circumferential Young's modulus by the three measured frequencies are, respectively, 135 kPa, 137 kPa, and 125 kPa. The estimated modulus is very consistent with the three resonant frequency measurements. The values of these three estimations are well within the range of arterial elastic modulus from published papers. With this method, the estimation of the Young's modulus of the artery only requires the diameter of the artery, but doesn't need the thickness of the artery which is difficult to measure with accuracy and precision.

Clinical studies report that impaired endothelial function is associated with Cardio-Vascular Diseases (CVD) and their risk factors. One commonly used mean for assessing endothelial function is Flow-Mediated Dilation (FMD). Classically, FMD is quantified using local indexes e.g. maximum peak dilation. Although such parameters have been successfully linked to CVD risk factors and other clinical variables, this description does not consider all the information contained in the complete vasodilation curve. Moreover, the relation between flow impulse and the vessel vasodilation response to this stimulus, although not clearly known, seems to be important and is not taken into account in the majority of studies. In this paper we propose a novel global parameterization for the vasodilation and the flow curves of a FMD test. This parameterization uses Principal Component Analysis (PCA) to describe independently and jointly the variability of flow and FMD curves. These curves are obtained using computerized techniques (based on edge detection and image registration, respectively) to analyze the ultrasound image sequences. The global description obtained through PCA yields a detailed characterization of the morphology of such curves allowing the extraction of intuitive quantitative information of the vasodilation process and its interplay with flow changes. This parameterization is consistent with traditional measurements and, in a database of 177 subjects, seems to correlate more strongly (and with more clinical parameters) than classical measures to CVD risk factors and clinical parameters such as LDL- and HDL-Cholesterol.

Stroke arrives without warning 80% of the time, leading to death and disability in a large number of the 700,000 it strikes in the USA each year. After discussing the special characteristics of screening instruments and the particular challenges of imaging the carotid artery, we report on our progress in developing an instrument to screen for the carotid stenoses that cause the majority of these strokes.

Wall shear rate (WSR) is the derivative of blood velocity with respect to vessel radius at the endothelial surface. The product of WSR and blood viscosity is the wall shear stress (WSS) that must remain relatively high to maintain normal endothelial cell function, arterial health and prevent plaque formation. Accurate WSR estimation requires the lowest possible variance and bias for blood velocity estimates near the wall. This situation is achieved for conditions where the echo signal-to-noise ratio (eSNR) and spatial resolution for velocity are high. We transmitted coded pulses, i.e., those with time-bandwidth product greater than 1, to increase eSNR from weak blood scatter without increasing instantaneous power or reducing spatial resolution. This paper is a summary of WSR measurements from a flow phantom where a variety of acoustic pulses were transmitted: frequencymodulated (FM) codes and phase-modulated (PM) codes were compared with uncoded broadband and narrow band pulse transmissions. Both simulation and experimental results show that coded-pulse excitation increases accuracy and precision in WSR estimation when compared to standard pulsing techniques. Additionally, PM codes can reduce WSR errors more than FM codes for equal pulse energy. This reduction in WSR error could greatly extend the application of ultrasound in the study of cardiovascular disease.

The intima-media thickness (IMT) of the carotid artery is an important biomarker for the clinical prognosis and diagnosis of atherosclerosis and stroke. This paper presents a new approach, pixel compounding, to enhance the resolution of the intima-media vascular layers in ultrasound B-scan images and provide increased image resolution for a more precise measurement. First, homomorphic transformation is used to estimate the lumped point spread function (PSF) of the images, then, the images are deblurred with the estimated PSF, and finally, a non-homogeneous anisotropic diffusion algorithm is used to further enhance the resolution of the image. The homogeneous part of the algorithm is used to suppress speckle while enhancing the coherent structures, specifically the edges; the non-homogeneous part (likelihood estimator) progressively adds the details from succeeding frames in the image sequence for an optimal and sub-pixel resolved image. Phantom studies have shown 300% improvement on Peak Distance Standard Deviation and nearly 100% improvement on Average Half Peak Width, indicating significant resolution enhancement.

Thermoacoustic tomography (TAT) is an emerging imaging technique with great potential for a wide range of biomedical imaging applications. It is customary in TAT to assume that the object is acoustically homogeneous, which can result in image artifacts in medical applications. In this work, we investigate an iterative reconstruction approach for TAT that can compensate for acoustic heterogeneities via inversion of a generalized Radon transform imaging model. We demonstrate numerically that the generalized Radon transform model can be inverted uniquely and stably by use of only half of the acquired measurement data. The effects of imperfect knowledge of the acoustic heterogeneity map are also investigated.

The paper presents a new ultrasonic attenuation imaging method which might be used as a new imaging modality, targeted at breast cancer diagnostics. Two approaches based on ultrasonic imaging are combined together, namely the estimation of ultrasound attenuation coefficients from pulse-echo B-mode imaging data and an ultrasound computer tomography imaging technique. A recently published method for estimation of the ultrasound attenuation coefficient using the log--spectrum analysis is applied to radiofrequency signals acquired by an ultrasound computer tomography system to estimate images of the attenuation coefficients. The examined volume (e.g. female breast) is enclosed by several thousand ultrasound transducers. Radiofrequency signals from all transducers using all sending positions are recorded. Compared to the known ultrasound attenuation tomography methods, not only the directly transmitted signal, but also the reflected and scattered signals are processed here, i.e. substantially more information is used. The method is presented in its initial stage. The applied algorithm is derived using simplifying assumptions which will be relaxed in further research. However, even at this stage the resulting attenuation image is of higher quality than the standard attenuation imaging methods applied to the same data set.

Ultrasound computer tomography is an imaging method capable of producing volume images with high spatial resolution. The imaged object is enclosed by a cylindrical array of transducers. While one transducer emits a spherical wavefront (pulse), all other transducers are recording the radiofrequency (RF) a-scans simultaneously. Then another transducer acts as the emitter and so on. In this paper we describe the image reconstruction method and an enhanced algorithm for the a-scan preprocessing. The image reconstruction is based on a 'full aperture sum-and-delay' algorithm evaluating the reflected and scattered signals in the a-scans. The a-scans are modelled as the tissue response of the imaged object convoluted with the shape of the ultrasound pulse, which is determined by the transfer function of the transducers and the excitation. Spiking deconvolution and blind deconvolution with different parameters are used to build inverse filters of the ultrasound pulse. Applying the inverse filters to the a-scans results in sharper signals which are used for image reconstruction. Smallest scatterers of 0.1 mm size corresponding to one fifth of the used ultrasound wavelength are visible in the reconstructed images. Compared to conventional b-scans the resulting images show an approximately tenfold better resolution.

Acoustic diffraction tomography (DT) is an inversion scheme that can reconstruct the spatially variant acoustic properties of a scattering object. In this work, we develop and investigate a novel reconstruction algorithm for reconstructing separate images of the density and compressibility fluctuations of nonviscoelastic scattering objects. The reconstruction algorithm is derived by identifying and exploiting tomographic symmetries and the rotational invariance of the problem. The proposed reconstruction algorithm is implemented numerically and demonstrated by use of computer-simulation studies.

We report preliminary results from our investigation of in vivo prostate elastography. Fewer than 50% of all prostate cancers are typically visible in current clinical imaging modalities. Elastography displays a map of strain that results when tissue is externally compressed. Thus, elastography is ideal for imaging prostate cancers because they are generally stiffer than the surrounding tissue and stiffer regions usually exhibit lower strain in elastograms. In our study, digital radio-frequency (RF) ultrasound echo data were acquired from prostate-cancer patients undergoing brachytherapy. Seed placement is guided by a transrectal ultrasound (TRUS) probe, which is held in a mechanical fixture. The probe can be moved in XYZ directions and tilted. The probe face, in contact with the rectal wall, is used to apply a compression force to the immediately adjacent prostate. We also used a water-filled (acoustic) coupling balloon to compress the prostate by increasing the water volume inside the balloon. In each scan plane (transverse), we acquired RF data from successive scans at the scanner frame rate as the deformation force on the rectal wall was continuously increased. We computed strain using 1D RF cross-correlation analysis. The compression method based on fixture displacement produced low-noise elastograms that beautifully displayed the prostate architecture and emphasized stiff areas. Balloon-based compression also produced low-noise elastograms. Initial results demonstrate that elastography may be useful in the detection and evaluation of prostate cancers, occult in conventional imaging modalities.

Conventional ultrasound systems acquire an image by scanning the environment successively line by line. This limits the frame rate. Furthermore the used conventional signal affects the resolution and the penetration depth. In this contribution we propose to use modified mutually orthonormal Hermite series in a coded excitation system yielding a full image in only one transmission. Therefore a higher frame rate can be achieved. Additional properties of these codes offer the possibility to increase the resolution and the penetration depth. Preliminary results show the effectiveness of the proposed approach.

This paper presents an automatic tissue segmentation methodology for High-Resolution Ultrasonic Transmission Tomography (HUTT) imagery of biological organs. This method combines a recent segmentation approach: the L-level set active contours algorithm with unsupervised clustering using the agglomerative hierarchical k-means algorithm. The active contours algorithm has been recently explored as a powerful tool for image segmentation since it automatically decomposes a given image into 2L segment classes by utilizing L level set functions and finding the optimal boundaries of the 2L segment classes so that the pixel feature values of each segment are as homogeneous as possible. Unfortunately, the algorithm is often trapped at local minima due to the intrinsic non-convexity of the cost function, especially for noisy data. To overcome this problem, we introduce a multi-stage multi-resolution analysis that optimizes the active contours at successive resolutions of the image data. The resulting segments are then re-clustered by subsequent agglomerative hierarchical k-means clustering that seeks the optimal clusters yielding the minimum within-cluster distance in the feature space. The preliminary studies reported here indicate that this proposed methodology can enhance the accuracy of soft tissue segmentation and provide fully automatic tissue differentiation without any user intervention except for specifying the number of level set functions L.

Doppler has been used for many years for cardiovascular exploration in order to visualize the vessels walls and anatomical or functional diseases. The use of ultrasound contrast agents makes it possible to improve ultrasonic information. Nonlinear ultrasound imaging highlights the detection of these agents within an organ and hence is a powerful technique to image perfusion of an organ in real-time. The visualization of flow and perfusion provides important information for the diagnosis of various diseases as well as for the detection of tumors. However, the images are buried in noise, the speckle, inherent in the image formation. Furthermore at portal phase, there is often an absence of clear contrast between lesions and surrounding tissues because the organ is filled with agents. In this context, we propose a new method of automatic liver lesions segmentation in nonlinear imaging sequences for the quantification of perfusion. Our method of segmentation is divided into two stages. Initially, we developed an anisotropic diffusion step which raised the structural characteristics to eliminate the speckle. Then, a fuzzy competitive clustering process allowed us to delineate liver lesions. This method has been used to detect focal hepatic lesions (metastasis, nodular hyperplasia, adenoma). Compared to medical expert’s report obtained on 15 varied lesions, the automatic segmentation allows us to identify and delineate focal liver lesions during the portal phase which high accuracy. Our results show that this method improves markedly the recognition of focal hepatic lesions and opens the way for future precise quantification of contrast enhancement.

Left-ventricular (LV) segmentation is essential in the early detection of heart disease, where left-ventricular wall motion is being tracked in order to detect ischemia. In this paper, a new method for automated segmentation of the left-ventricular chamber is described. An autocorrelation-based technique isolates the LV cavity from the myocardial wall on 2-D slices of 3D short-axis echocardiograms. A morphological closing function and median filtering are used to generate a uniform border. The proposed segmentation technique is designed to be used in identifying the endocardial border and estimating the motion of the endocardial wall over a cardiac cycle. To this purpose, the proposed technique is particularly successful in border delineation by tracing around structures like papillary muscles and the mitral valve, which constitute the typical obstacle in LV segmentation techniques. The results using this new technique are compared to the manual detection results in short-axis views obtained at the papillary muscle level from 3D datasets in human and canine experiments in vivo. Qualitatively, the automatically-detected borders are highly comparable to the manually-detected borders enclosing regions in the left-ventricular cavity with a relative error within the range of 4.2% - 6%. The new technique constitutes, thus, a robust segmentation method for automated segmentation of endocardial borders and suitable for wall motion tracking for automated detection of ischemia.

Time domain algorithms that solve the Khokhlov--Zabolotzskaya--Kuznetsov (KZK) equation are described and implemented. This equation represents the propagation of finite amplitude sound beams in a homogenous thermoviscous fluid for axisymmetric and fully three dimensional geometries. In the numerical solution each of the terms is considered separately and the numerical methods are compared with known solutions. First and second order operator splitting are used to combine the separate terms in the KZK equation and their convergence is examined.

Strain imaging is useful for visualizing information related to tissue stiffness. However, strain is a parameter that depends on the boundary conditions, tissue connectivity and geometry. As a result, tissue hardness cannot be quantitatively evaluated from the strain distribution. Therefore, reconstruction of the elastic modulus (Young's Modulus) distribution has been investigated for quantitative evaluation of tissue hardness. A method has been recently proposed [NITT 00] to calculate locally the Young's modulus of tissues from the estimations of 3D displacement field within the medium. This approach requires a specific annular ultrasonic probe. The aim of our work, based on Nitta's approach, is to build a Young modulus mapping using clinical ultrasonic equipment. Results from finite-element simulations and a physical phantom are presented.

In this paper a method for spatio-temporal encoding is presented for synthetic transmit aperture ultrasound imaging (STA). The purpose is to excite several transmitters at the same time in order to transmit more acoustic energy in every single transmission. When increasing the transmitted acoustic energy, the signal to noise ratio will increase. However, to focus the data properly using the STA approach, the transmitters have to be separated from each other. This is done by dividing the available spectrum into several subbands with a small overlap. Separating different transmitters can be done by bandpass filtering. Therefore, the separation can be done instantaneously without the need for further transmissions, unlike spatial encoding relying on Hadamard or Golay coding schemes, where several transmissions have to be made before the decoding can be done. Motion artifacts from the decoding can, thus, be avoided. To further increase the transmitted energy, the excitation waveforms are designed as linear frequency modulated (FM) signals. This makes it possible to maintain the full excitation amplitude during most of the transmission. The design of the separation filters will also be discussed. The method was tested using the experimental ultrasound scanner RASMUS and evaluated using a reference setup with a linear FM excitation waveform and STA beamforming. The point spread function (PSF) was measured on a wire phantom in water. A wire phantom with an attenuating medium was also measured, where the proposed method achieved approximately 2 cm improvement in penetration depth. The signal to noise ratio was also measured, where the gain was approx. 7 dB in comparison to the reference.

Currently synthetic aperture flow methods can find the correct velocity magnitude, when the flow direction is known. To make a fully automatic system, the direction should also be estimated. Such an approach has been suggested by Jensen (2004) based on a search of the highest cross-correlation as a function of velocity and angle. This paper presents an experimental investigation of this velocity angle estimation method based on a set of synthetic aperture flow data measured using our RASMUS experimental ultrasound system. The measurements are performed for flow angles of 60, 75, and 90 deg. with respect to the axial direction, and for constant velocities with a peak of 0.1 m/s and 0.2 m/s. The implemented synthetic aperture imaging method uses virtual point sources in front of the transducer, and recursive imaging is used to increase the data rate. A 128 element linear array transducer is used for the experiments, and the emitted pulse is a 20 micro sec. chirp, linearly sweeping frequencies from approximately 3.5 to 10.5 MHz. The flow angle could be estimated with an average bias up to 5.0 deg., and a average standard deviation between 0.2 deg. and 5.2 deg. Using the angle estimates, the velocity magnitudes were estimated with average standard deviations no higher than 6.5% relative to the peak velocity.

In conventional ultrasound color flow mode imaging, a large number (~500) of pulses have to be emitted in order to form a complete velocity map. This lowers the frame-rate and temporal resolution. A method for color flow imaging in which a few (~10) pulses have to be emitted to form a complete velocity image is presented. The method is based on using a plane wave excitation with temporal encoding to compensate for the decreased SNR, resulting from the lack of focusing. The temporal encoding is done with a linear frequency modulated signal. To decrease lateral sidelobes, a Tukey window is used as apodization on the transmitting aperture. The data are beamformed along the direction of the flow, and the velocity is found by 1-D cross correlation of these data. First the method is evaluated in simulations using the Field II program. Secondly, the method is evaluated using the experimental scanner RASMUS and a 7 MHz linear array transducer, which scans a circulating flowrig. The velocity of the blood mimicking fluid in the flowrig is constant and parabolic, and the center of the scanned area is situated at a depth of 40 mm. A CFM image of the blood flow in the flowrig is estimated from two pulse emissions. At the axial center line of the CFM image, the velocity is estimated over the vessel with a mean relative standard deviation of 2.64% and a mean relative bias of 6.91%. At an axial line 5 mm to the right of the center of the CFM image, the velocity is estimated over the vessel with a relative standard deviation of 0.84% and a relative bias of 5.74%. Finally the method is tested on the common carotid artery of a healthy 33-year-old male.

Fast visualisation of cerebral microcirculation supports diagnosis of acute stroke. However, the commonly used CT/MRI-based methods are time consuming, costly and not applicable to every patient. The bolus perfusion harmonic imaging (BHI) method is an ultrasound imaging technique which makes use of the fact, that ultrasound contrast agents unlike biological tissues resonate at harmonic frequencies. Exploiting this effect, the contrast between perfused and non-perfused areas can be improved. Thus, BHI overcomes the low signal-to-noise ratio of transcranial ultrasound and the high impedance of the skull. By analysing image sequences, visualising the qualitative characteristics of an US contrast agent bolus injection becomes possible. The analysis consists of calculating four perfusion-related parameters, Local Peak Intensity, Time To Peak, Area Under Curve, and Average Rising, from the time/intensity curve and providing them as colour-coded images. For calculating these parameters the fundamental assumption is that image intensity corresponds to contrast agent concentration which in turn shows the perfusion of the corresponding brain region. In a clinical study on patients suffering from acute ischemic stroke it is shown that some of the parameters correlate significantly to the infarction area. Thus, BHI becomes a less time-consuming and inexpensive bedside method for diagnosis of cerebral perfusion deficits.

When an ultrasonic examination is performed, a segmentation tool would often be a very useful tool, either for the detection of pathologies, the early diagnosis of cancer, the follow-up of the lesions, ... Such a tool must be both reliable and accurate. However, because of the relatively poor quality of ultrasound images due to the speckled texture, the segmentation of ultrasound data is a difficult task. We have previously proposed to tackle the problem using a multiresolution bayesian region-based algorithm. Such an approach, applied to very noisy images, leads to good segmentation results. For computation time purposes, a multiresolution version has been implemented. In order to improve the quality of the segmentation, we propose to get more information about the properties of the tissues and take it into account during the segmentation process. Some acoustical parameters have thus been computed, either directly from the images or from the Radio-Frequency (RF) signal. The parameters used are the Integrated BackScatter (IBS), the density of scatterers, and the Mean Central Frequency, which is a measurement related to the attenuation of ultrasound waves in the media. To test the influence of the acoustical parameters in the segmentation process, a set of numerical phantoms has been computed using the Field software. Each phantom consists in two regions with different acoustical properties : the density of scatterers and the scattering amplitude. From both the simulated RF signal and images, parameters have been computed and segmentation has been processed for each phantom. The quantification of the segmentation quality is based on the number of correctly classified pixels and it has been computed for all the combinations of acoustical parameters. Segmentation results performed on agar-gelatine phantoms with different inclusions are also presented and illustrate the interest of a multiparametric segmentation approach.

Utilization of tools during surgical interventions sets the problem of their accurate localization within biological tissue. The ultrasound imaging represents an inexpensive and a flexible approach for a real-time image acquisition of tissue structure with metal instruments. There are several difficulties involving processing of ultrasound images: Their noisy nature makes the localization task difficult; the objects appear irregular and incomplete. Our task is to determine the position of a curvilinear electrode in biological tissue from a three-dimensional ultrasound image. Initially, the data are segmented by thresholding and processed with the randomized version of the RANSAC (R-RANSAC) algorithm. The curvilinear electrode is modeled by a three-dimensional cubic curve. Its shape is subject to check using a curvature measure in the hypothesis evaluation step of the R-RANSAC algorithm. Subsequently, we perform the least squares curve fitting to the data that have been marked by the R-RANSAC as the ones corresponding to the sought object. The position estimation is optimal with respect to the mean square criterion. Finally, the localization of the electrode tips is carried out by a hypothesis testing on the distances between projections of inliers on the estimated curve. The algorithm has been tested on real three-dimensional ultrasound images of a tissue mimicking phantom with a curvilinear object. From the results, we conclude that the method is very stable even if the data contain high percentage of outliers. The computational cost of the algorithm shows the possibility of real-time data processing.

We have developed a breast ultrasound image retrieval system for the decision support of classification of breast lesions on ultrasound. Biopsy-proved direct-digital 1,034 data were used to construct the main DB engine.The image database is constructed by about 1,034 cases (malignant cases are about 400. Benign cases are about 600.) And all cases were biopsy-proven. Two kinds of retrieval images of full size image and ROI image can be executed. Computer-generated malignant likelihood is calculated based on the retrieved images. So, user can infer what's the basis of computer's malignancy likelihood

The swept scan technique has been employed in high frequency ultrasonic imaging with a mechanically scanned single element transducer. It improves the scanning speed over the discrete scan technique. Although the technique has been widely adopted, effects of continuous scanning on accuracy of blood flow estimation have not been studied. In this paper, we use the 2-D spatial spectrum (i.e., k-space) to describe such effects on both axial and lateral velocity estimations. The proposed k-space modeling on 2-D motion exhibits the following characteristics. Spatial spectrum of an axially moving object is the shifted version of that of a stationary object. The shifted amount is directly related to the axial velocity and increases as the axial spatial frequency increases. The lateral spatial spectrum bandwidth is proportional to the relative lateral motion between the transducer and the object. Based on these properties, an algorithm for 2-D vector velocity estimation is proposed. Both simulations and flow phantom experiments were conducted. A 45-MHz transducer was used and the transducer was scanned at 20 mm/s. The Doppler angles ranged from 29° to 90° and the flow velocity was between 15 and 30 mm/s. Results show that the proposed k-space velocity estimator had an average angle estimation error of 2.6° and standard deviations from 2.2° to 8.2°. When the swept scan effects were ignored, the average angle estimation error was 14.2° and standard deviations from 1.4° to 3°. Thus, benefits of the proposed k-space vector velocity estimator in high frequency ultrasonic imaging were clearly demonstrated.

The characterization of the endothelial function is one of the most attractive research topics in modern vascular medicine. The evaluation of the flow-mediated vasodilation (FMD) of the brachial artery is a widely used measurement technique. Despite its widespread use, this technique has some limitations due to the difficulties in obtaining an accurate measurement of such a small vessel (3 to 5 mm) by using ultrasounds. The system we present in this paper can automatically measure the diameter of the artery with high accuracy on each image of a video sequence. Furthermore, it processes the data in real-time, thus providing the physician with an immediate response while the examination is still in progress. The main part of the system is a video processing board based on a state-of-the-art digital signal processor (DSP). The board acquires the video signal generated by the ultrasound equipment which furnishes a longitudinal section of the artery vessel. For each image, the DSP automatically locates the two borders of the vessel and subsequently computes the diameter. The algorithm used to automatically locate the borders of the vessel is based on a new operator of edge detection which was derived from the first absolute central moment. Tests in many clinical centers proved that the system provides very accurate measurements and is a remarkable step forward toward a more systematic evaluation of the FMD.

Aim: The objective of this work is to determine the optimal basis function to perform wavelet analysis for tissue characterization of radio frequency intravascular ultrasound (IVUS) backscattered data. This is the most important step in wavelet analysis as it ensures accurate decomposition of the original signal into the various frequency bands. The criterion to choose the mother wavelet that is best suited to the data depends on the intended application. Wavelet families possessing properties like orthogonality, regularity, stability and admissibility have previously been shown to have application in tissue characterization. Algorithm: Depending on the usable data bandwidth known from previous studies we decomposed data using a 4-level decomposition scheme. We then calculated Shannon’s entropy for every level and employed “minimum Shannon entropy criterion” to determine the best mother wavelet for signal decomposition. According to this criterion, accurate decomposition is indicated when the total entropy of the daughter (decomposed) levels is lower than the entropy of the parent level. Analysis and Results: We acquired 40 MHz IVUS data ex-vivo from 10 left anterior descending (LAD) coronary arteries. Data was acquired such that each frame comprised of 256 scanlines. Next, we randomly selected 3 scanlines for each LAD and applied the above-mentioned Shannon entropy criterion for these 30 scanlines. We analyzed 23 mother wavelets from different families. Daubechies 3rd order wavelet accurately decomposes 29/30 scanlines at all levels. Daubechies 6th order wavelet appears optimal for 21/30 scanlines. Future direction: To obtain more precise signal decomposition, the optimal mother wavelet should be selected at every decomposition level. The best mother wavelet is indicated by the lowest Shannon entropy for that particular level.

Vibro-acoustography is an imaging method based on the vibro-acoustic response of the object to a low-frequency radiation force of ultrasound. Here, we present the results of a study on detection mass lesions by vibro-acoustography. Experiments were conducted on excised human liver tissues that included focal mass lesions which were a few mm to a few cm in diameter. The 3 MHz transducer used for this purpose provided a 0.7 mm lateral resolution. The focal length of the transducer was long enough to cover the entire 5 mm thickness of the specimen. Several scans of each sample were obtained. Resulting images distinctively showed the normal liver tissue and the mass lesions. Masses appeared with enhanced boundary and rough textures in VA which allowed us to delineate the masses from the surrounded tissue. These results suggest that vibro-acoustography may be a clinically useful imaging modality for detection of mass lesions in soft tissue.

Elasticity imaging has great potential in soft tissue characterization since the tissue elasticity is usually related to some abnormal, pathological process. Internal tissue deformation induced by externally applied mechanical forces has been evaluated to characterize tissue elasticity. For a quantitative elasticity imaging, material parameter such as Young's modulus must be reconstructed from ultrasonic measurement of internal displacement. A method to estimate the elastic modulus of an isotropic, inhomogeneous, incompressible elastic 3-D medium using measured displacement data is formulated by inversely solving the forward problem for static deformation. A finite-element based model for static deformation is proposed and then rearranged for solving the distribution of the shear modulus of the soft tissue from a knowledge of the displacement within the tissue. When the force boundary condition is unknown, it reconstructs the relative value of the elastic modulus of the tissue using the displacement data. The feasibility of the proposed method is demonstrated using the simulated deformation data of the simple three-dimensional inclusion problem. The performance of the algorithm with noise in the diplacement measurement data is teseted using numerical simulations. The results show that the relative shear modulus may be reconstructed from the displacement data measured locally in the region of interest within an isotropic, incompressible medium, and that the relative shear modulus can be recovered to some degree of accuracy from only one-dimensional displacement data. Details of how to apply this method under clinical conditions is also discussed.

We present an open-source MATLAB toolkit for ultrasound calibration. It has a convenient graphical user interface which sits on top of an extensive API. Calibration using three different phantoms is explicitly supported: the cross-wire phantom, the single-wall phantom, and the Hopkins phantom. Image processing of the Hopkins phantom is automated by making use of techniques from binary morphology, radon transform and RANSAC. Numerous calibration and termination parameters are exposed. It is also modular, allowing one to apply the system to original phantoms by writing a minimum of new code.

An etalon is an optical resonator where light is confined to a thin transparent layer having reflecting coatings on the sides. Optical reflection from this structure is highly sensitive to local mechanical perturbation. This is the basic principle allowing these devices to act as 2D ultrasound detector arrays. Optical probing of the etalon surface defines the array geometry and detection size of each element in the array. Element size on the order of several microns is easily realized. The detection bandwidth is limited primarily by the acoustic propagation time thru the layer thickness. We have developed etalon structures optimized for high frequency ultrasound detection using thin polymer layers (less than 10 μm). The detection bandwidth of these devices is typically 100MHz. The sensitivity of the etalon detector was demonstrated to be comparable to that of a piezoceramic detector. The etalon was integrated into a photoacoustic imaging system. High resolution images of phantom targets and biological tissue (nerve cord) were obtained. The additional information of optical absorption obtained by photoacoustic imaging, along with the high resolution detection of the etalon, offer unique advantages for intravascular and neurological imaging applications.