Official Proceedings - AIUM

American Institute of Ultrasound in Medicine Proceedings J Ultrasound Med 32(suppl):S1–S134, 2013deconvolution between the hydrophone output voltage and the hydrophonefrequency-dependent complex sensitivity. We have previouslyreported a method for measuring the magnitude and phase of hydrophonesensitivity using time delay spectrometry (TDS). The goal of this work isto assess the improvement in the accuracy of estimates of acoustic outputparameters (pulse intensity integral and peak rarefactional pressure) usingcomplex deconvolution.Methods—In the first set of experiments, a swept-frequencyTDS system was used to measure magnitude and phase responses of severaltypes of hydrophones used in medical ultrasound exposimetry. Theseincluded polyvinylidene difluoride spot-poled membrane, needle, and capsuledesigns. Measurements were performed using 4 broadband sourcetransducers to measure hydrophone sensitivity over the band from 1 to 30MHz. In the second set of experiments, 6 hydrophones were used to measurethe acoustic pressure waveform generated by a 3-MHz single-elementsource transducer. The voltage waveforms acquired in the second set of experimentswere deconvolved with sensitivities measured in the first set ofexperiments. The effect of deconvolution on measurements of the pulse intensityintegral and peak rarefactional pressure was measured.Results—The effect of deconvolution on measurements ofpulse the intensity integral and peak rarefactional pressure sometimes exceeded10%.Conclusions—The frequency dependence of hydrophone sensitivitycan have a substantial impact on measurements of the pulse intensityintegral and peak rarefactional pressure. In these cases, complexdeconvolution can be used to compensate for frequency-dependent hydrophonesensitivity.1538679 Pulmonary Hemorrhage Induced by Diagnostic UltrasoundRevealed by Growth of Comet Tail Artifacts in the ImageDouglas Miller Radiology, University of Michigan, AnnArbor, Michigan USAObjectives—Ultrasound examination of the lung has becomean important part of chest medicine, particularly for point-of-care diagnosisin emergency rooms and intensive care units. The objective of thisstudy was to explore the potential for lung injury, which may arise fromthe interaction of ultrasound pulses with alveolar gas, using a rat model ofpulmonary diagnostic ultrasound.Methods—Anesthetized rats were prepared by shaving theright thorax and then mounted in a 37°C water bath. A linear array (CL15-7, HDI 5000; Philips Healthcare, Andover MA) was used for B-mode imagingof the right lung at ≈7.6 MHz. A low mechanical index (MI) of 0.21was used to align the scan plane through an intercostal space. The MI thenwas raised for 5 minutes to higher settings in different groups of 5 rats. Fora sham group, the rats were prepared but not scanned. The real-time imagewas recorded and evaluated for occurrence of comet tail artifacts (CTAs),which are indicative of alveolar fluid. The lungs were evaluated for the sizeof any pulmonary hemorrhages (PHs).Results—For the highest available MI (0.9), the image immediatelydisplayed growing CTAs, which rapidly spread across the entirebright-line image of the lung surface. The CTAs appeared within secondsat MI = 0.7 or 0.9 but more slowly at lower MIs. Contusion-like PHs werefound on the lungs, which appeared to have a one-to-one correspondencewith the CTAs in the image. The proportion of positive results was statisticallysignificant for MI = 0.52 (4 of 5 rats; P < .01) but not for MI = 0.37(2 of 5, P > .1), relative to no PH in shams.Conclusions—PH was induced in a rat model of pulmonary diagnosticultrasound at moderate MIs, and this bioeffect was indicated by thegrowth of CTAs in the image. The induction of PHs by pulsed ultrasoundwas discovered over 20 years ago but appeared to pose little risk to patients,because only incidental scanning of the lung was expected. However,direct scanning, which occurs for pulmonary applications, may carry a riskof pulmonary injury for some patients. More information will be needed toprovide safety guidance consistent with optimal diagnostic imaging.1528109 Evaluation of Definity Stability Over Time Using DoublePassive Cavitation DetectionMarianne Gauthier, 1,2 * Daniel King, 1,3 William O’Brien Jr 1,21Bioacoustics Research Laboratory, 2 Electrical and ComputerEngineering, 3 Mechanical Science and Engineering, Universityof Illinois at Urbana-Champaign, Urbana, Illinois USAObjectives—Definity is the first ultrasound contrast agent(UCA) approved by the US Food and Drug Administration that offersflexible dosing and administration through intravenous bolus injection orcontinuous intravenous infusion. In a clinical context (for diagnosis, therapy,and bioeffect studies), temporal stability of the UCA can be criticalusing either infusion or bolus: infusion implies stability of the microbubblesduring the time of the injection, while bolus may be repeated to acquireseveral images for the same patient, implying the microbubbles toexhibit the same properties over time.Methods—This study’s aim was to assess the stability of Definityover time. Experiments were performed using the double passive cavitationdetection (DPCD) method, allowing the evaluation of the collapsethresholds of an isolated microbubble based on the detection of postexcitationsignals occurring 1 to 5 microseconds after the principle excitationof the bubble. Five sets of DPCD experiments (3-cycle tone bursts at thecentral frequency of 2.8 MHz) were performed over 3 weeks. For eachset of experiments, 5% and 50% collapse thresholds were determined withtheir 95% confidence interval (CI) based on the generalized linear modelregression performed using MatLab. We also compared the size distributionof each tested microbubble set.Results—Statistical analysis exhibited no significant differencesin the bubble size distributions and the 5% and 50% collapse thresholdsmeasured using the DPCD method (Table 1).Conclusions—Definity microbubbles have been found to bestable over the 3 weeks of experiments from the size distribution and the5% and 50% collapse thresholds points of view. Definity can be used withoutextra precaution concerning its temporal stability. (Supported by NationalInstitutes of Health grant R37EB002641.)Table 1. Bubble Diameter, 5% and 50% Postexcitation Thresholds ± 95% CIsEvaluated Over 3 Weeks5% Postexcitation 50% PostexcitationBubble Diameter Threshold ± 95% CI, Threshold ± 95% CI,Group ± 95% CI, μm MPa MPa1 1.40 (1.28–1.52) 0.022 (0.001–0.191) 0.173 (0.001–0.559)2 1.23 (1.12–1.34) 0.05 (0.001–0.277) 0.455 (0.005–0.944)3 1.39 (1.26–1.52) 0.077 (0.003–0.226) 0.38 (0.075–0.671)4 1.26 (1.13–1.39) 0.116 (0.015–0.269) 0.464 (0.162–0.72)5 1.42 (1.31–1.53) 0.058 (0.001–0.235) 0.315 (0.01–0.663)1541018 Arrival Time Estimation in a Sparsely Sampled HemisphericTransducer ArrayJason Tillett, 1 * Jeffrey Astheimer, 1 Robert Waag 1,2 1 Electricaland Computer Engineering, 2 Imaging Sciences, University ofRochester, Rochester, New York USAObjectives—Estimate waveform arrival time fluctuationscaused by propagation through a breast model in a sparsely sampledfaceted approximation of a hemispheric transducer array.Methods—A 3D pseudospectral k-space method was used tocalculate acoustic propagation from a point source located near the centerof an array of widely separated transducers. The point source, with a centerfrequency of 5 MHz and –6-dB bandwidth of 2.5 MHz, was situatednear the chest wall of a numeric anthropomorphic breast model, and thetransducer array surrounded the pendant boundary of the breast. The hemispherewas approximated using 40 triangular facets. The separation of elementsaveraged about 1.5 times the wavelength at 5 MHz, ie, about 3S42