Search

The derivation and theoretical evaluation of new wideband maximum likelihood strategies for the estimation of blood velocity using acoustic signals are presented. A model for the received signal from blood scatterers, using a train of short wideband pulses, is described. Evaluation of the autocorrelation of the signal based on this model shows that the magnitude, periodicity and phase of the autocorrelation are affected by the mean scatterer velocity, and the presence of a velocity spread target. New velocity estimators are then derived that exploit the effect of the scatterer velocity on both the signal delay and the shift in frequency. The wideband range spread estimator is derived using a statistical model of the target. Based on the point target assumption, a simpler wideband maximum likelihood estimator is also obtained. These new estimation strategies are analyzed for their local and global performance. Evaluation of the Cramer-Rao bound shows that the bound on the estimator variance is reduced using these estimators, in comparison with narrowband strategies. In order to study global accuracy, the expected estimator output is evaluated, and it is determined that the width of the mainlobe is reduced. In addition, it is shown that the height of subsidiary velocity peaks is reduced through the use of these new estimators.

Part I of the work presented the derivation and theoretical evaluation of new wideband maximum likelihood strategies for the estimation of blood velocity using acoustic signals. In Part II, the signal models and performance of the estimation strategies described in Part I are tested with experimental ultrasonic data. The ultrasonic data analyzed in the work verifies the theoretical model and predicted performance. The averaged correlation, verified experimentally, confirms that the correlation envelope can be used to estimate the velocity of scatterers, and that the shape of the correlation function conveys information regarding the velocity profile within tht sample volume. For both the wideband point and range spread estifl!lators, the predicted improvement in velocity resolution and the redu(tion in the height of subsidiary velocity peaks are demonstrated. Through the use of these estimation strategies, information regarding the mean velocity and velocity variation are available for each spatial location within the vessel. This information is presented using a three-dimensional (3-D) spatial velocity profile display, which appears to offer a number of advantages in the rapid identification of pathology.

The derivation and theoretical evaluation of new wideband maximum likelihood strategies for the estimation of blood velocity using acoustic signals are presented. A model for the received signal from blood scatterers, using a train of short wideband pulses, is described. Evaluation of the autocorrelation of the signal based on this model shows that the magnitude, periodicity and phase of the autocorrelation are affected by the mean scatterer velocity, and the presence of a velocity spread target. New velocity estimators are then derived that exploit the effect of the scatterer velocity on both the signal delay and the shift in frequency. The wideband range spread estimator is derived using a statistical model of the target. Based on the point target assumption, a simpler wideband maximum likelihood estimator is also obtained. These new estimation strategies are analyzed for their local and global performance. Evaluation of the Cramer-Rao bound shows that the bound on the estimator variance is reduced using these estimators, in comparison with narrowband strategies. In order to study global accuracy, the expected estimator output is evaluated, and it is determined that the width of the mainlobe is reduced. In addition, it is shown that the height of subsidiary velocity peaks is reduced through the use of these new estimators.

Part I of the work presented the derivation and theoretical evaluation of new wideband maximum likelihood strategies for the estimation of blood velocity using acoustic signals. In Part II, the signal models and performance of the estimation strategies described in Part I are tested with experimental ultrasonic data. The ultrasonic data analyzed in the work verifies the theoretical model and predicted performance. The averaged correlation, verified experimentally, confirms that the correlation envelope can be used to estimate the velocity of scatterers, and that the shape of the correlation function conveys information regarding the velocity profile within tht sample volume. For both the wideband point and range spread estifl!lators, the predicted improvement in velocity resolution and the redu(tion in the height of subsidiary velocity peaks are demonstrated. Through the use of these estimation strategies, information regarding the mean velocity and velocity variation are available for each spatial location within the vessel. This information is presented using a three-dimensional (3-D) spatial velocity profile display, which appears to offer a number of advantages in the rapid identification of pathology.

High-speed optical experiments demonstrate that the behavior of a polymer-shelled microbubble contrast agent in response to an acoustic pulse is qualitatively and quantitatively different from that of a lipid-shelled agent. The lipid-shelled agent expands in response to a two-cycle pulse, and at pressures approaching 1 MPa, both the shell and its contents fragment. The polymer-shelled agent remains largely intact at pressures up to 1.5 MPa and exhibits a different destruction mechanism: the polymer shell does not oscillate significantly in response to ultrasound; instead, a gas bubble is extruded and ejected through a shell defect while the shell appears to remain largely intact. (C) 2004 American Institute of Physics.

High-speed optical experiments demonstrate that the behavior of a polymer-shelled microbubble contrast agent in response to an acoustic pulse is qualitatively and quantitatively different from that of a lipid-shelled agent. The lipid-shelled agent expands in response to a two-cycle pulse, and at pressures approaching 1 MPa, both the shell and its contents fragment. The polymer-shelled agent remains largely intact at pressures up to 1.5 MPa and exhibits a different destruction mechanism: the polymer shell does not oscillate significantly in response to ultrasound; instead, a gas bubble is extruded and ejected through a shell defect while the shell appears to remain largely intact. (C) 2004 American Institute of Physics.