Get e-book Acoustical Imaging

Free download. Book file PDF easily for everyone and every device. You can download and read online Acoustical Imaging file PDF Book only if you are registered here. And also you can download or read online all Book PDF file that related with Acoustical Imaging book. Happy reading Acoustical Imaging Bookeveryone. Download file Free Book PDF Acoustical Imaging at Complete PDF Library. This Book have some digital formats such us :paperbook, ebook, kindle, epub, fb2 and another formats. Here is The CompletePDF Book Library. It's free to register here to get Book file PDF Acoustical Imaging Pocket Guide.

Introduces the gauge invariance property of acoustic equation of motion, with applications in the elastic constants of isotropic solids, time reversal acoustics, negative refraction, double negative acoustical metamaterial and acoustical cloaking. Contains up to date treatments on latest theories of sound propagation in random media, including statistical treatment and chaos theory. Includes a chapter devoted to new acoustics based on metamaterials, a field founded by the author, including a new theory of elasticity and new theory of sound propagation in solids and fluids and tremendous potential in several novel applications.

Covers the hot topics on acoustical imaging including time reversal acoustics, negative refraction and acoustical cloaking. Get A Copy. Hardcover , pages. Published July 23rd by Wiley first published April 24th More Details Other Editions 6. Friend Reviews. To see what your friends thought of this book, please sign up.

To ask other readers questions about Acoustical Imaging , please sign up. Lists with This Book. This book is not yet featured on Listopia. Community Reviews. Showing Rating details. All Languages. More filters. Sort order. Kaiser rated it it was amazing Aug 04, Alessandrini, A. Palladini, L. Fujiwara, K. Okubo, and N. Tagawa Medical Ultrasound Image Deconvolution. Shin, R. Prager, H. Gomersall, N. Kingsbury, G. Treece, and A. Burov, A. Shmelev, and O. Rumyantseva Epilogue. Lee Author Index.

Borup Techniscan Medical Systems, Inc. Callahan Techniscan Medical Systems, Inc. KG, Bensheim, Germany N. Johnson Techniscan Medical Systems, Inc. Kobayashi Honda Electronics Co. Ltd, Toyohashi, Japan M. Sakai Kanebo Cosmetics Inc. Smith Techniscan Medical Systems, Inc. Wiskin Techniscan Medical Systems, Inc. Ledgerwood Abstract Our computer-aided diagnostic CADx tool uses advanced image processing and artificial intelligence to analyze findings on breast sonography images.

The goal is to standardize reporting of such findings using well-defined descriptors and to improve accuracy and reproducibility of interpretation of breast ultrasound by radiologists. This study examined several factors that may impact accuracy and reproducibility of the CADx software, which proved to be highly accurate and stabile over several operating conditions. Ordinarily, if a lesion appears solid or indeterminate, biopsy is frequently recommended, although current estimates of the False Negative FN rate and Positive Predictive Value PPV for breast ultrasound are not definitive.

Furthermore, even with combined information from mammography and ultrasound it is apparent that each radiologist may apply a different decision threshold to recommend biopsy of a suspicious mass. In addition, operator variability and image quality are often identified as important issues in medical breast ultrasound US , perhaps one of the more difficult medical imaging procedures to perform [2]. BC began as a research tool, its clinical utility and ease of use have evolved considerably. The cases in the Reference Library have known findings confirmed by biopsy, aspiration or 2-year benign follow-up and include a detailed BI-RADS report with descriptions and impressions provided by consensus of expert radiologists.

BC may also be used to create individual teaching files of interesting cases. The BC CADx is the first computer-aided diagnostic system as compared to a computer-aided detection tool that is traditionally designed for independently identifying suspicious areas on a screening examination. In the CADx domain, some level of suspicion may already exist from prior imaging results or physical exams palpation, for example , which are the most common indications for breast sonography.

BC is tailored as an aid to the radiologist for interpretation of the diagnostic breast ultrasound examination but the underlying technology is applicable to other medical imaging procedures including breast MRI, mammography, computed tomography of the lung, carotid artery flow analysis, etc.

The purpose of this study was to examine specific factors that may impact the accuracy and reproducibility of results from the developed CADx software in future clinical use and to summarize its overall performance.

Course description

When a suitable case was found by the Lead Radiologist where truth was confirmed Method to Standardize Breast Ultrasound Interpretation 5 2-year benign follow up or biopsy , all ultrasound images in the study were examined to ensure they were free of graphic overlays or markers, had at least two views of each mass, had minimal artifacts, had conclusive pathology results for biopsies, had a corresponding mammogram available, etc.

The mix of case findings in the database was: simple cysts , The age range was 21—90 years old. The case mix of findings and age distribution of patients match very closely 2. BC uses case-based reasoning analysis derived from known findings of the most similar cases in the retrieved cluster, which are determined based on measurement of parameters in the following categories: margins, shape, echogenicity, echo texture, orientation, and posterior acoustic attenuation pattern. BC requires no classifier training unlike artificial neural networks commonly used in computer-aided detection systems for screening examinations.

Processing is nearly instantaneous. The mass is dark, with some internal echoes and posterior Fig. The CLA score 2. The margin of the mass is segmented from the surrounding tissue and depicted by the isocontour. Seven cases are automatically retrieved and displayed with contours on the right listed in rank order of Relative Similarity to the unknown mass. These results are very similar to those achieved in our prior studies and to those reported elsewhere for breast sonography. The results of the radiologist reader study, both ROC performance and inter-reader variability, offer a relevant benchmark for tests of BC CADx performance.

In this laboratory, not clinical, setting AZ for BC 0. Agreement between the image pairs as measured by the kappa statistic was 0. No significant difference was found when radial and anti-radial images of the same mass were compared or when images of the same mass were compared from two different examinations on two different scanners. These results suggest that the BC CADx is reproducible and independent of scanner or operator factors, although this issue warrants additional study.

Method to Standardize Breast Ultrasound Interpretation 9 We appreciate that the radiologist is faced with a decision-making task during interpretation of breast ultrasound that is influenced by many factors, not all of which were included in this study. Nonetheless, the improved results represent a very high performance. They also suggest that the BC CADx approach may be successful in the aims of aiding radiologists to reduce biopsies on benign masses and to achieve higher Specificity with minimal impact on Sensitivity. Investigation of the BC CADx in a clinical investigational setting is under way where a group of experienced radiologists are reading the database of cases twice: first in the normal clinical manner following the ACR BI-RADS protocol, and second while using the electronic reporting system of the BC CADx including consideration of the CLA score prior to making their final recommendation.

An interval of approximately 6 months will occur between imaging review sessions. This new study examines how the use of BC may impact accuracy of reading performance of the radiologists and it will document any changes in reader variability. We gratefully acknowledge the support of Julie Phan, B. References 1. Stavros, A. Radiology , — 2. Lee, H. Acoustical Imaging 28, — 4. Galperin, M. Acoustical Imaging 28, — 5. Acoustical Imaging 29, — 6. Dorfman, D. Metz, C. Lazarus E. Severin Abstract Biometrics is a rapidly evolving scientific and applied discipline that studies possible ways of personal identification by means of unique biological characteristics.

Such identification is important in various situations requiring restricted access to certain areas, information and personal data and for cases of medical emergencies. A number of automated biometric techniques have been developed, including fingerprint, hand shape, eye and facial recognition, thermographic imaging, etc.

All these techniques differ in the recognizable parameters, usability, accuracy and cost. Among these, fingerprint recognition stands alone since a very large database of fingerprints has already been acquired. Also, fingerprints are key evidence left at a crime scene and can be used to indentify suspects. Therefore, of all automated biometric techniques, especially in the field of law enforcement, fingerprint identification seems to be the most promising.

We introduce a newer development of the ultrasonic fingerprint imaging. The proposed method obtains a scan only once and then varies the C-scan gate position and width to visualize acoustic reflections from any appropriate depth inside the skin. Also, B-scans and A-scans can be recreated from any position using such data array, which gives the control over the visualization options. By setting the C-scan gate deeper inside the skin, distribution of the sweat pores which are located along the ridges can be easily visualized.

This paper discusses different setups, acoustic parameters of the system, signal and image processing options and possible ways of 3-dimentional visualization that could be used as a recognizable characteristic in biometric identification. Maev et al. Such identification is critical in situations requiring restricted access to areas, data and objects or in case of some medical emergencies. Direct measurement is supposed to be more foolproof and protected in comparison with classical linking some kind of tag ID, key or password to the person.

There are several techniques for biometrics authentication which are actively developing and promoting by different companies [1]. The oldest one is the face recognition which was performed manually during centuries. Hand shape recognition also is not reliable enough due to various changes in the dimensions and placement of the original measurements of the hand. Identification based on scanning of iris and retinal patterns is much more accurate but requires complicated positioning and can be affected by the contact lens. All of the above methods provide data which are not compatible with existing law enforcement databases and information accumulated in criminal practice.

Such compatibility exits for fingerprint biometrics. Widespread and long-term usage of this method gives essential experience and familiarity. Classical pen and ink are gradually replaced by more convenient and sophisticated devices utilizing the optical, capacitance and ultrasonic techniques [1—3]. Each of ten easily assessable unique fingerprints has relatively small area, what makes the scanning process fast and convenient.

This study extends scanning acoustic microscope technology for fingerprint imaging. The method is insensitive for surface contamination and allows use not only the surface grooves pattern, but also the internal structure of the fingers. Mapping of elements of this structure, such as sweat pores and scars, increase information and add security measurements. A specialized holder was designed to keep finger of volunteer motionless during this time. Finger surface was pressed against acoustically transparent polystyrene plate with thickness 2 mm Fig. Acoustic gel was used between the plate and finger for Ultrasonic Method for 3D Fingerprint Characteristics 13 acoustic lens acoustic reflections XY scanning water coupling plate finger skin plate gel or water coupling Fig.

The whole assembly was submerged into water tank under the lens scanning plane Fig. These signals obtained during one scanning stroke were compiled into two- dimensional B-Scan which resembles a 0. Series of B-scans were further complied and stored into 3D matrix of information. A horizontal slice of this volume at chosen depth can be plotted as acoustic C-scan. This technique allowed us to only once Fig.

C-scan Paper print Fig. Also, B-scans and A-scans can be recreated from any point. Therefore, this gives the total control over the visualization options. The pattern is obviously identical; however the grooves in the acoustical image light and the optical one are inverted dark. The white spots located along the ridges of dermis at acoustical image are sweat pores. Given that their distribution should be unique for each individual, this provides additional means of personal identification which is not affected by any changes accidental or intentional of the finger surface conditions.

Focusing on close undersurface layer reveals their position as well as depth and parameters of dermal layers Fig. This structure is consistent even in presence of mechanical distortions, like a scar Fig. The collected information can be directly compared to the very large already existing database of fingerprints. The method allows for not only the fingerprint pattern to be analyzed but also other dermis elements such as scars and sweat pores. These details are as unique as the print and unlike the surface cannot be surgically altered or removed.

The dermal layers are visible in the B-scan along with the bottom of the plate C-scan Paper print Fig. The same technique can be applied to characterize the fingernail and its underlying tissue and characteristics which are also unique from one individual to another. Design of specialized cylindrical multi-transducer scanner will make the method fast and convenient. The large area of the finger nail to nail 2. The digitally 16 R. The only downside is the use of a medium such as water or gel.

Millard, K. Carnahan Conf. Security Technol. Schneider, J. Schmitt, R. Laugier Abstract Current histological methods can miss micrometastases E. Feleppa et al. Typically, nodes in the region considered to be draining the site of the tumor are removed surgically, cut into thick sections or blocks of 2—3 mm in depth, fixed and embedded. For some cancers, e. In the touch-prep procedure, dissected nodes are cut in half and the cut surfaces are pressed onto microscope slides for immediate cytological examination of cells that remain attached to the slide surface.

In the case of the axillary nodes of breast-cancer patients, three or four nodes may be identified by the dye and removed for examination, which spares the great majority of the nodes in the axillary bed. If sentinel-node metastases are detected at that time, then a formal dissection can be performed immediately. If metastases are not detected at that time, but are detected after a subsequent complete histological procedure, then the patient needs to be recalled to undergo a formal dissection. Sentinel-node procedures are attractive for many patients because, if the histology proves to be negative, then the often-severe side effects of a formal lymphadenectomy can be avoided if no metastases in fact exist.

Unfortunately, these procedures are prone to producing false negatives. Frozensection and touch-prep procedures leave virtually the entire node unexamined, and easily can miss small metastases that are not located on or near the plane of the cut. Furthermore, typical full histological procedures only evaluate the thin sections taken from the surfaces of the 2- to 3-mm-thick blocks, which allow the small metastases lying between the cut surfaces to remain undetected. Therefore, a need exists for a method of evaluating the entire node and identifying suspicious regions that warrant more-detailed histology by the pathologist immediately after node dissection.

If sufficiently reliable, the method also could be used to evaluate sentinel nodes upon dissection and to determine immediately whether a formal dissection is warranted. The remainder of this article describes a method based on spectrum analysis of radio frequency RF , high-frequency-ultrasound HFU echo signals that initially has demonstrated a very encouraging potential for discriminating cancerous from non-cancerous tissue in lymph nodes and for providing an urgently needed method to guide pathology, and, in particular, for supporting decisions regarding formal Detection of Metastases in Dissected Lymph Nodes 19 node dissections in sentinel-node procedures.

Spectrum-analysis methods for tissue typing have been described in numerous publications [1—4], but are summarized very briefly in the Methods section below. Clinical tasks, i.

Submission history

Data and results were transferred among research sites via the internet. Dissected nodes were placed in isotonic saline for transport to the pathology laboratory. In the pathology laboratory, individual nodes were selected and excess perinodal tissue was removed from each specimen. Nodes were pinned individually to a sound-absorbing pad by inserting pins through the thin layer of remaining fat.

Pinned nodes were placed in a water bath filled with phosphatebuffered isotonic saline and scanned at room temperature for acquisition of RF echo signals as described below. In the case of sentinel nodes, each half node was pinned, immersed, and scanned. After echo-signal data were acquired, nodes were inked, as shown in Fig. The red of Fig. Histological sections were obtained in the horizontal planes depicted in Fig. The transducer position was adjusted in the z direction to place the focus in the center of the node or 2 mm deep within the node, whichever was less in terms of depth into the node.

Scanning was performed at room temperature. In Fig. The scanning apparatus is shown in Fig. Figure 2a shows the complete apparatus, including the positioning stages; Fig. Data have been acquired from nodes of patients with colorectal, breast, and stomach cancers.

The most-common type of cancer in our data set was colorectal cancer; the data include approximately nodes from approximately 90 colorectal-cancer patients. Accordingly, results described here were computed for 38 nodes dissected from 22 colorectal-cancer patients. Ultrasound echo-signal data for calibration, i. This allows the transducer to be immersed in water, which obviates the necessity to correct for often severely temperature-dependent propagation properties of oils. Having a weakly reflecting interface of known reflectivity allowed the settings of the data-acquisition system e.

Instead, all fixed nodes were cut in half, embedded, and sectioned in planes parallel to the cut surfaces. Stained and mounted thin sections then were examined by a pathologist under high-power magnification. After high-power 22 E. The second step was applying QUS methods based on spectrum analysis to compute QUS parameters such as spectral slope and intercept and estimates of scatterer size and acoustic concentration [1—3].

Acoustic concentration was defined by Lizzi as the number concentration times the square of the relative acoustic impedance of the scatterers compared to their surroundings [3]. A region-based semiautomatic 3D segmentation method was used. It involved a watershed-transform of the 3D gradient of the low-pass filtered, down sampled by a factor of 8 , log-compressed envelope of the 3D RF data [5]. The watershed regions were classified as saline, fat or tissue using a maximum-likelihood classifier based on the mean backscatter of each region.

The maximum-likelihood classifier took into account transducer diffraction and then employed depth-dependent thresholds to classify the regions [5]. Finally, minor artifacts of the 3D segmentation were corrected by an expert using custom software. Spectrum analysis occurred within the segmented nodal tissue using cylindrical regions of interest ROIs that were 1 mm in diameter and length.

Spectralparameter values and scatterer-property estimates were computed as the ensemble average of all scan-vector segments in the ROI. Normalized spectra were computed for every scan vector in the ROI using the method described by Lizzi [1—3]; scatterer-property estimates applied a Gaussian form factor using the method described by Insana [6].

Spectral slope values and scatterer size estimates were corrected for measured attenuation in the perinodal fat layer and for an assumed attenuation coefficient of 0. Spectral amplitudes and parameter values were expressed in dBr, i. An interactive GUI was developed that displayed the node volume as 3 orthogonal planes in the 3D volume of the node. Mouse movement controlled the position of a line cursor in any one of the 3 views, which controlled the display of the orthogonal plane corresponding to the line cursor. A histogram displays the intercept-value distribution for each node 24 E.

Negative and positive nodes of a colorectal-cancer patient are shown in Fig. The upper node is cancer free and the lower one contains cancer. The images display color-encoded spectral-intercept values along with the semi-automatically segmented saline, fibro-adipose, and nodal-tissue regions in each of 3 orthogonal planes. The displayed planes are selected by the cursors in either of the other pair of planes. The intercept value is depicted in grey scale with lighter values indicating intermediate values and darker indicating the greater deviations, lower more-negative or higher more-positive values.

A calibrated grey bar is displayed between the nodes. Histograms depict the parameter-value distribution within the node. Classifier performance was expressed as the area under the ROC curve [7]. Linear discriminant analysis was performed to determine the classification performance possible using multiple variables and to determine objectively which variables contributed most to the classification [8].

To do this, the variables that visual inspection of means and standard deviations suggested might have value for discriminating between cancerous and noncancerous tissue were input into SPSS SPSS, Chicago, IL ; the selected variables were scatterer size, acoustic concentration, spectral slope, and spectral intercept.

Furthermore, spectral-intercept values and scatterer-size estimates for positive and negative values were input individually into ROCKIT software to derive classifier-assessing ROC curves directly. SPSS did not compute standard errors in the area estimates. Cancerous tissue in the nodes tended to have larger scatterer sizes as well as the more-positive intercept values shown in these figures. Interestingly, the current results and those of a decade ago indicate that the attenuation-independent intercept parameter possibly can serve as a cancer-identifying parameter without requiring segmentation or attenuation correction.

The results described here also are consistent with results published earlier for 4 nodes of an individual colorectal-cancer patient segmented using 2D rather than the current 3D approach; 10 in those data, 1 node was positive and 3 were negative for metastases, and the positive node exhibited larger scatterer-size estimates and higher intercept values. The current results offer the hope that results derived from a larger number of nodes, particularly cancer-containing nodes, will lead to an effective method of assessing dissected nodes.

These assessments will enable targeted pathology evaluations of nodes in traditional histological procedures and will enable more-rapid decision making in sentinel-node procedures. In both applications, a reduction in false-negative determinations and improved staging of a wide variety of cancers will be valuable clinical benefits. In the short term, future studies will emphasize processing of a greater number of positive nodes and thereby obtain a better representation of metastatic tissue in lymph nodes. These studies also will seek to automate the crucial segmentation algorithms and to automate them as much as possible.

Classification methods will be evaluated by spatially comparing tissue-typing results with cancerous and non-cancerous regions demarcated by the pathologists and represented in 3D in our 26 E. This evaluation effort will require careful co-registration of histological and ultrasonic data using non-rigid deformation techniques to correct for distortions and shrinkage occurring during histological preparation.

The evaluations will be crucial in validating the application of spectral properties associated with nodefilling metastases to those of smaller, sometimes multiple foci within nodes — the micrometastases that are the object of this research. In the long term, future studies will emphasize reducing the processing to clinically more-useful and practical means of imaging suspicious tissue in 3D.

Acoustical Imaging by Walter Arnold, Sigrun Hirsekorn -

A reduction of processing time and the development of automated analysis and imaging tools will be essential for these methods to be acceptable and practical clinically. Although only 38 nodes from 22 patients have been evaluated to date, and only 7 of those nodes contained a significant cancer volume, areas under the ROC curves equaling 1. Lizzi, F. Feleppa, E. Ultrasound Med. Cancer Biomark. Coron, A. Insana, M. Investigative Radiology 21, — 8. McLachlan, G. Wiley, New York, NY 9. Tateishi, T. Greenleaf Abstract The cardiovascular diseases atherosclerosis, coronary artery disease, hypertension and heart failure have been related to stiffening of vessels and myocardium.

Noninvasive measurements of mechanical properties of cardiovascular tissue would facilitate detection and treatment of disease in early stages, thus reducing mortality and possibly reducing cost of treatment. While techniques capable of measuring tissue elasticity have been reported, the knowledge of both elasticity and viscosity is necessary to fully characterize mechanical properties of soft tissues.

In this article, we summarize the Shearwave Dispersion Ultrasound Vibrometry SDUV method developed by our group and report on advances made in characterizing stiffness of large vessels and myocardium.

IN4010 – Acoustic Imaging

The method uses radiation force to excite shear waves in soft tissue and pulse echo ultrasound to measure the motion. The speed of propagation of shear waves at different frequencies is used to generate dispersions curves for excised porcine left-ventricular free-wall myocardium and carotid arteries. An antisymmetric Lamb wave model was fitted to the LV myocardium dispersion curves to obtain elasticity and viscosity moduli. The results suggest that the speed of shear wave propagation in four orthogonal directions on the surface of the excised myocardium is similar.

These studies show that the SDUV method has potential for clinical application in noninvasive quantification of elasticity and viscosity of vessels and myocardium. These diseases have been related to stiffening of the vessels and myocardium.

Decreased elasticity stiffening of the myocardium leads I. Nenadich et al. Current clinical tools used for evaluation of diastolic function rely on measurements of ejection fraction using ultrasound Doppler imaging. Ejection fraction coefficient is not a reliable indicator of diastolic function since patients with normal ejection fractions and preserved systolic function often present with angina, edema and other symptoms of heart failure [2]. Decreased arterial elasticity stiffening can lead to hypertension and, according to Dolan, et al.

Changes in mechanical properties of soft tissue due to pathological conditions have been known for centuries. In recent decades, significant effort has been directed towards producing a technique capable of non-invasive characterization of tissue elasticity. These techniques are mostly based on inducing tissue deformation and detecting the response with either magnetic resonance imaging or ultrasound [4, 5]. Unlike ultrasound techniques, MRI is not limited by focal length and acoustic window, but is fairly expensive and therefore unlikely to become widely available.

Thus, the field of ultrasound elastography has developed and offers various methods of characterizing tissue viscoelasticity. Early techniques [6—9] were capable of producing relative tissue stiffness maps, but such information was sufficient only when the diagnosis of the diseases where the pathologic and healthy tissues present at the same time for comparison. Diffuse disorders such as liver fibrosis, left-ventricular free wall stiffening, and atherosclerosis affect the elasticity of the entire tissue and their diagnoses require methods capable of quantifying tissue stiffness.

Shear wave propagation velocity has been used to quantify tissue stiffness, [10—12] largely because the shear moduli from body tissues such as muscle, liver, dermis, cartilage and bone differ by several orders of magnitude [12]. Shear wave attenuation in soft tissue is fairly high which allows for high spatial resolution because of the use of high frequency waves. These techniques, however, neglect tissue viscosity and therefore underestimate the elasticity by ignoring the complex part of shear modulus.

A liver fibrosis study by Huwart, et al. Thus, a technique capable of quantifying both tissue elasticity and viscosity would be of great benefit in clinical settings.

  • New & Forthcoming Titles.
  • Acoustical Imaging: Techniques and Applications for Engineers.
  • Vintage Wisconsin Gardens: A History of Home Gardening?
  • Haematin Enzymes. A Symposium of the International Union of Biochemistry Organized by the Australian Academy of Science Canberra;

Here, we describe a non-invasive ultrasound technique developed by our group that uses shear wave velocity dispersion to quantify both elasticity and viscosity of soft tissue [14, 15]. This technique uses harmonic radiation force to induce deformation and pulse-echo ultrasound to detect shear wave motion. As previously reported by our group, in vivo measurements of liver viscoelastic moduli using SDUV are in good agreement with the values reported by other groups [15]. In addition to describing SDUV technique, we summarize the progress we have made in quantifying material properties of LV myocardium and large vessels.

Due to narrow beamwidth and constant force along the beam axis z-axis near the focus the push transducer produces cylindrical shear waves [14]. Equation 1 shows that for large r according to simulations, larger than one tenth of the shear wavelength the phase delay and distance from the excitation are linearly related. An alternative approach is to move the excitation point and keep the detect transducer in one location. As mentioned above, a pulse echo ultrasound transducer is used to detect the motion. Ultrasound pulses are transmitted at the location of interest at a pulse repetition rate of few kilohertz.

Each point in the time domain of the returning echo signal corresponds to a specific region of the tissue along the beam axis. Cross-spectral correlation of the echoes is used to calculate tissue displacement as a function of time [16]. A specialized Kalman filter is applied to the displacement versus time data to extract the motion at the excitation frequency of the push transducer and estimate the shear wave amplitude and phase [17].

In commercial scanners it is possible to steer the ultrasound beam. This can be used for the steering of the pushing and the detection beam. The pushing beam is applied at a specific site in the tissue to generate a shear wave that propagates through it. After the excitation, the system changes to a pulse echo mode, where the detection echo beam is steered to at least 2 different positions in the area adjacent to the excitation to obtain phase differences for the propagating shear wave.

As shown in Fig. Tr is the period of the pushing toneburst and Tb is the duration. Tprf is the period of the tracking sequence, the transmission of the pulse echo and the reception of the pulse echo. This figure is modified from Urban, et al. After each of these push tonebursts, the system switches to the detection mode, and the pulse echo signal for tracking the motion is generated with a frequency fprf of 1—4 kHz, depending on the application.

The motion of the tissue is extracted from the echo signal using cross spectrum correlation. Due to the repetition frequency of the pushing beam, the shear waves generated contain pure-tone frequencies harmonics of the repetition frequency up to 1 kHz, and all this can be recorded simultaneously. Therefore the shear wave velocity dispersion needed to characterize the viscoelastic properties of tissues can be easily obtained with one excitation sequence.

This reduces the time of acquisition and the heating of the tissue, making it feasible to explore different points of the tissue during the same exam. The tissue sample was embedded in gelatin inside a container and mounted on a stand to provide stability. The gelatin was contained to the edges of the sample in order to eliminate coupling. In this set of experiments, the push transducer was replaced by a mechanical actuator to ensure large motion.

The experimental set up was otherwise as in Fig. The detect transducer was used to track motion at multiple locations in four orthogonal directions on the surface plane. Kalman filter and Equation 3 were used to estimate the speed at different frequencies. Shear wave speed at different frequencies dispersion measured in four orthogonal directions are shown in Fig.

The phase values of the particle displacement were fairly constant throughout the thickness z-axis which is characteristic for anti-symmetric Lamb wave motion. The experimental data were fitted with a Lamb wave mathematical model solid lines proposed by Kanai [19]. The measured shear wave speeds agree with previously reported values [19].

To our best knowledge viscoelastic coefficients are not well established in literature. These results suggest that the Lamb wave model can be used to fit the experimental results and that the porcine LV myocardium elasticity and viscosity are similar in four orthogonal directions. In the future, we intend to perform in vivo open chest studies in pigs to examine the capabilities of this method. Figure 4, shows the motion recorded on the wall of an excised pig carotid after exciting it with radiation force.

The artery was mounted on a metallic frame, covered with gelatin to simulate the surrounding tissue of in vivo carotids, 34 I. Experimental data are shown as circles connected with lines and Lamb wave model as solid lines and pressurized to 50 mmHg using a column of water. The total time of pushing sequence was 80 ms. Panel C shows the difference in the frequency components of the two signals presented in panel A and B.

Notice how the magnitude of fast Fourier transform FFT of the SDUV signal has amplitude only at harmonics of the repetition frequency, 50, , , etc, up to 1 kHz, giving a better signal to noise ratio at the harmonics of the repetition frequency. In this experiment, the motion of the wall was measure at 41 different points spaced 0. Cross correlation of the motion was performed to find the time shifts between consecutive distances. The time shift versus distance was fitted with a linear regression and the group velocity extracted from the slope of the fitted line.

In the case of the pig carotid pressurize to 50 mmHg the group velocity was 8. Panel C shows the magnitude of the FFT of the motion of the wall when excited with the impulse or with the SDUV pushing sequence Figure 5 shows the shear wave dispersion curve for the same carotid, pressurized to 70 mmHg. Currently, our lab is working on a model to fit the dispersion data to get values of the elasticity and the viscosity of the vessels similar to the one described above for the myocardium.

Wave speed is estimated by fitting linear regression through the phase data at multiple points. Shear wave dispersion measurements are used to estimate tissue elasticity and viscosity. SDUV technique has been successfully used to estimate liver viscoelasticity in pigs in vivo and has potential for application in estimating elasticity and viscosity in vessels and myocardium. American Heart Association. Dallas, Texas: American Heart Association; Accessed on March 17, Gary, R.

Heart Lung. Dolan, E. Hypertension 47, — 4. Gao, L. Greenleaf, J. Lerner, R. Ophir, J. Imaging 13 2 , — 8. Fatemi, M. Science , 82—85 9. Nightingale, K. Sandrin, L. IEEE Trans. Control 49 4 , — Sarvazyan, A. Huwart, L. NMR Biomed. Chen, S. J Acoust Soc Am. Hasegawa, H. Zheng, Y. Urban, M. Kanai, H. Tortoli Abstract Traditional Doppler methods only measure the axial component of the velocity vector.

The lack of information on the beam-to-flow Doppler angle creates an ambiguity which can lead to large errors in velocity magnitude estimates. An original approach was recently introduced, in which two ultrasound beams with known relative orientation are directed towards the same vessel, one being committed to perform a Doppler measurement, while the second beam has the specific task of detecting the beam-to-flow angle. In this paper, an angle-tracking procedure allowing the Doppler angle to be automatically determined with high accuracy is presented.

The procedure is based on the real-time estimation of suitable Doppler spectrum parameters obtained from an M-line associated to a sub-aperture of a linear array probe. Such parameters are used to steer the M-line towards a direction corresponding to a desired beam-flow angle. Knowledge of this angle is finally exploited to obtain the velocity magnitude through the classic Doppler equation related to the second beam.

The implementation of the method on a new ultrasound machine and its validation through in vitro and in vivo tests are reported. Dallai et al. The standard approach to obtain angle-independent blood velocity estimates is based on the combination of Doppler measurements taken along multiple US beams intersecting in the region of interest [1]. The same sample volume SV is insonified by two or more probes whose beam axes are oriented along directions describing a known inter-beam angle. The Doppler equations related to the frequencies obtained by the different probes, are then trigonometrically combined to provide an estimate of both the velocity magnitude and the flow direction.

A method employing the classic dual-beam configuration in an original way has been recently introduced [2]. The flow direction is identified by forcing the reference probe to be transversely oriented to the flow itself. This condition is known to produce Doppler spectra having unique features of symmetry around zero frequency. Once the flow direction has been identified, the second beam can be used to estimate the velocity magnitude through an angle-corrected Doppler frequency measurement. Although the technique has been thoroughly validated in vitro, for its practical in vivo application it is useful that the needed transverse beam-flow angle is automatically achieved.

In this paper, we present an automatic angle tracking method, based on the real-time computation of suitable Doppler spectral parameters. The method has been implemented in the programmable ULtrasound Advanced Open Platform ULA-OP , an experimental system connected to a linear array probe, which was recently developed in our lab [3]. In vitro experiments conducted in both steady and pulsatile flow conditions are reported, showing that the flow direction can be identified with small errors whichever the initial Doppler angle is over a wide range.

In that particular direction, in fact, the Doppler spectrum of the backscattered echo is substantially symmetrical around the zero frequency. Since the mean Doppler frequency of the received signal is proportional to the cosine of the nominal Doppler angle, even a small deviation from the desired transverse orientation causes a visible loss of symmetry, and the SSI correspondingly drops down. The procedure starts by fixing, within a B-mode image, the position of the Doppler SV to be investigated.

The spectrum of echo data received from the SV is monitored in a real-time spectrogram, and used to continuously provide the spectral symmetry index, SSI, and the spectral mean frequency, fd.

Acoustical Imaging

The reference M-line steering angle is automatically modified in order to decrement fd and to increment the SSI. The steering correction is initially made in coarse steps, which are progressively decreased while the SSI increases. Once the SSI has become larger than a suitable threshold, the steering angle is no more modified, and the flow direction is assumed to be perpendicular to the current reference M-line.

The corresponding spectral data are finally converted to velocity magnitude through the Doppler Equation 1. The echo-signals are channeled through low noise amplifiers, sampled by bit 50 Msps Analog-to-Digital Converters, and dynamically beamformed according to the desired steering angle and apodization factor. The ULA-OP system was programmed in such a way that two reference and measuring, respectively M-lines are simultaneously controlled, i. This is feasible thanks to the DSP firmware architecture that permits concurrent processing modules to coexist and run without hampering each other.

Each module is connected to a different data stream and employs a distinct algorithm, therefore when activated the processing module that encapsulates the tracking algorithm is connected to the samples gathered by the reference M-line. For blood flow observation, other Vector Doppler Based on Automatic Transverse Angle Tracking 43 modules, such as the one needed for spectrogram computation, can be allocated and connected to the measuring M-line.

The tracking algorithm continuously evaluates the spectrum of the signal backscattered along the reference M-line throughout point FFTs. In order to increase stability, up to consecutive and partially overlapped Doppler spectra are averaged to extract more robust spectral parameters. The optimal steering direction is determined according to the sign of fd. The linear array probe was positioned to produce longitudinal scans of the tube see Fig.

At the beginning of each experiment, the B-mode display of the ULAOP system was used to set the SV at a convenient location typically, in the center of the tube. The spectrograms obtained from the reference line was continuously monitored on the PC display and used to achieve the desired transverse beam-flow angle. Once the reference line had reached its final orientation, the real time spectrogram assumed a stable symmetrical behavior and the raw Doppler data related to the measuring line were acquired in a PC file to allow an off-line analysis.

The entire procedure was repeated several times, for a given volume flow, by changing each time the initial probe orientation. The repeatability of the method was assessed by calculating, for each in vitro volume flow, the coefficient of variation CV , i. Over 20 measurements made in steady flow conditions produced CVs between 1. Four groups of measurements, each corresponding to a different velocity profile, were obtained in pulsatile flow conditions. The velocity curves obtained from the acquisitions corresponding to the same pulsatile flow profile were also compared to each other. As an example, Fig.

The SD mean value, normalized with respect to the mean velocity, produced the CV for this group. CVs between 5. All measurements were made by the same operator. The SV was located in the center of the right common carotid artery, at least 2 cm from the bifurcation. After 44 A. The indicated angles were estimated through the described procedure removing the probe from the neck of the volunteer, the operator proceeded with the next acquisition. For each subject, 3 acquisitions were made, each including at least 3 cardiac cycles, depending on the heart rate. The peak systolic velocities measured in each acquisition were averaged to generate a mean value.

The mean values obtained from different measurements on the same volunteer see Fig. The measured velocities and the resulting CVs ranged between 0. The mean CV was 6. The preliminary experiments on volunteers suggest the method appropriate for in-vivo applications. In all cases where the desired transverse orientation is within the steering capabilities of the used linear array probe, the proposed method makes the proposed dual-beam technique suitable for clinical application in human vessels. The velocity profile estimated through the measuring line is shown on the bottom Acknowledgments The authors wish to thank all the staff of the Microelectronic Systems Design Laboratory, and in particular Stefano Ricci and Luca Bassi, for their great contribution to this work.

Dunmire, B. Tortoli, P. Bassi, L. Ricci, S. Akiyama Abstract We propose a new contrast-echo method using counter-crossed beams of two ultrasonic frequencies as an ultrasound diagnostic for cancer. Sum and difference frequency components derived from nonlinear vibration of the contrast agents microbubbles driven by dual-frequency ultrasound are used in the C-CBCE method.

In this study, we used Sonazoid microbubbles as we attempted to detect the sum frequency component generated by Sonazoid fixed in agar gel. We also measured the in-channel flow velocity of the Sonazoid. In the contrast-echo imaging method, all echo signals except those from microbubbles can be eliminated by using secondary signals generated by nonlinear vibration of the microbubbles contrast agents [4].

Thus, the distribution of the microvascular system is generally determined by a harmonic-imaging method that visualizes the second-harmonic component in the echo signals. However, a second-harmonic component is generated during propagation through biological soft tissues as well as by nonlinear vibration of the microbubbles. In particular, when measuring a slow blood flow velocity such as in the diagnosis of hepatocellular carcinoma, the influence of the propagation through biological soft tissues cannot be ignored.

Therefore, separation of the echoes caused by microbubbles in the blood flow from echoes received from biological soft tissues is an important issue. In order to address this issue, we propose a new contrast-echo imaging method using counter-crossed beams of two different ultrasonic frequencies. The flow of T. Eura et al. In this study, we attempted to detect the sum frequency component generated from microbubbles fixed in agar gel.

In addition, we measured the flow velocity of the microbubbles in the channel. Two transducers with different resonant frequencies face each other. A broad ultrasonic beam with a frequency of f1 is formed by a transducer. Another transducer forms a narrow ultrasonic beam with a frequency of f2. By extracting these frequency components from the received echo signal, we can measure very slow blood velocities associated only with the contrast echo.

The continuous transmission of ultrasound with frequency f1 from a fixed transducer and a conventional scan featuring repeated transmission of pulsed ultrasound waves with frequency f2 can be combined to visualize the spatial distribution of blood velocity. We used Sonazoid for the microbubbles. The experiment system is described below.

Two concave ceramic transducers face each other in degassed water in a water tank. One transducer is driven by 50 cycles of a sinusoidal wave at 2 MHz with sound pressure amplitude at the focal point of kPa. The other transducer is driven by 14 cycles of a sinusoidal wave at 2. The Sonazoid was fixed in agar gel and arranged in the crossed area.

The ultrasound wave was scattered by the Sonazoid in the agar gel and received by a needle-type hydrophone with a diameter of 1. We compared the result for the agar gel with microbubbles with that for the agar gel with particulates. Here, a Hamming window is used on the received signal before carrying out a FastFourier transform in order to eliminate the transient response of the transducer from the echo signal.

The amplitude is normalized by the fundamental component f2 , as seen in Fig. Only the fundamental components f1 and f2 are observed in Fig. Therefore, these harmonic components and the sum frequency component are generated by nonlinear vibration of the microbubbles in the area where the beams cross.

In order to explain the experiment setup in detail, the front view of a water tank is depicted in Fig. As seen in Fig. Power Amp. The cycle period of the input signals is 1 ms. A silicon rubber tube with an internal diameter of 4 mm is located at the focal point of the concave transducer, and Sonazoid flows into the tube with degassed water. The ultrasound wave is scattered by the Sonazoid in the tube and is received by a PVDF hydrophone near the concave transducer.

The received signals are analyzed as follows. Each scattered wave from the Sonazoid is put through a gate function and subjected to quadrature detection by the sum-frequency component 4. The scattered signals are arranged to produce a two-dimensional signal whose horizontal axis is time and whose vertical axis is repetition time. The cycle period of the input signals is 1 ms, and thus the repetition time interval is 1 ms. The two-dimensional signal is subjected to Fourier transforms in the direction of the repetition time axis.

We used a period of high intensity to calculate the centroid frequency. Finally, we determined this value as a Doppler shift frequency in order to calculate the flow velocity in the tube. Figure 3 presents a map of the spectrum vs. We also measured the flow velocity of microbubbles in a silicon rubber tube. In our result, we extracted the sum frequency component generated by the nonlinear vibration of microbubbles from the echo signals.

Akiyama, I. Tanaka, S. In: Enflo B. Leighton, T. Academic Press, London, pp. Young, F. McGraw-Hill Companies, London, pp. Watanabe, R. Johnson Abstract This paper discuss a fully 3D nonlinear algorithm that results in a 3D quantitative estimate of breast tissue characteristics and a refraction corrected reflection algorithm RFCR that utilizes these estimates. The data are obtained from a specially designed clinical ultrasound breast scanner and processed on the device. We discuss the data collection process, a fast solution to the forward problem and a concomitant fast inverse scattering solution for the imaging problem.

We show how the resulting 3D tissue map is used in a refraction corrected reflection algorithm. However, the attainment of an inversion or imaging algorithm that utilizes the full waveform and the inherent nonlinearity of the inversion process, in a potentially useful time frame, has been lacking. Furthermore, the inversion should take place on a computational engine that can accompany the data acquisition device, and be reasonably inexpensive, if the device is to be clinically useful as a self-contained device.

Specifically the minimization is based on the Ribiere-Polak version of nonlinear conjugate gradients [2], therefore it requires a fast way to calculate the gradient of F and the step length. These calculations are shown below. Wiskin et al. The inversion algorithm is based on a type of approximate factorization of the Helmholtz wave equation that leads to a form of the phase screen approach described in U.

The transmitter sends out a broadband chirp signal from 0. This signal passes through the water bath, into the breast and finally reaches the receiver array on the opposite side of the breast. This array consists of horizontal elements and 6 rows vertically. The columns are 0. The reflection data are collected on three separate transceivers with different focal lengths of 2.

The purpose of the variable focal lengths is to allow overlapping focal regions resulting in greater depth of focus than one transceiver alone. The bandwidth of the transceivers is 2—8 MHz. Data collection occurs with the patient lying prone on the examination table, with the breast hanging pendant through the hole in the table, into the water bath, and between the transducer arrays.

As the breast is lowered into the water bath, the magnet is attracted to the top of the retention rod Fig. We proceed from low frequencies to high frequencies to avoid local minima. We image at 0. Note that this functional involves all views, and all levels simultaneously, that is, it is a true 3D algorithm. The algorithm at the highest level is described in Wiskin, et al. The calculation of the step-length and the gradient are detailed below. First a time of flight algorithm is used to create a series of initial distribution for speed of sound and attenuation at each level. These initial estimates are used in a series of 2D inverse scattering algorithms to create a series of 2D inverse scattering images at 56 J.

These 2D images are then concatenated together to form a 3D volume. This 3D volume is the starting estimate for the full 3D inverse scattering algorithm. The 3D algorithm is required to account for energy that is refracted out of plane. The 2D-algorithm gives an anomalously high result for the attenuation estimate. This is obtained in the following manj ner. The attenuation images are used to adjust the amplitude of the energy along the computed ray.

A data driven back-projection algorithm is used to place the reflected energy at its place of origin, via ray tracing, at each of 60 views, which are then compounded. However, no quantitative tissue characteristics are recovered.

Electromagnetic Acoustic Imaging (EMA) Simplified

The attenuation images are thought to be less accurate, due to the poorer numerical conditioning of the attenuation inversion problem. Figure 3 shows a fibro-adenoma exhibiting characteristic high speed, low to medium attenuation, and little internal structure in the reflection mode. Further results from patient-volunteers, and discussion, are shown in the accompanying paper in these proceedings. Additionally, these images could be used in a clinical setting, the topic of the following paper in these proceedings.

Inverse Scattering Theory 59 Fig. The speed of sound and attenuation maps produced can be used to create a refraction corrected reflection image. The refraction correction improves spatial resolution by allowing large angle compounding, which reduces speckle. In: Akiyama, I. Acoustical Imaging, vol. Springer, Dordrecht 2. Wiskin, J.

Sobre Nosotros

Springer, Dordrecht 3. Bailin, D. Institute of Physics Publishing, Bristol 4. Johnson, et al. Born, M. Johnson Abstract We discuss the results obtained using the inversion and refraction corrected reflection RFCR algorithms described in the companion paper in these proceedings. We show images for three patients created with these algorithms, from data collected with our clinical device. We discuss their potential clinical relevance, and their relationship to other, more conventional imaging modalities.

There have been two major types of ultrasound data processed: transmission data, and reflection data, with standard ultrasound utilizing reflection data. However, since most inversion algorithms avoid the inherent nonlinearity of the inversion process, the images tend to lack quantitative accuracy.