Session: 1E CONTRAST AGENTS II Chair: K. Ferrara University of California-Davis 1E-1 10:30 a.m. PULSE INVERSION DOPPLER FOR BLOOD FLOW DETECTION IN THE MACRO- AND MICRO-VASCULATURE WITH ULTRASOUND CONTRAST AGENTS M. BRUCE* 1,S.JENSEN 2,M.AVERKIOU 2,andJ.POWERS 2, 1 University of Washington, and 2 Philips Ultrasound. Corresponding e-mail: mbruce@u.washington.edu Visualization of macro-vessels is difficult using existing contrast agent imaging methods without blooming or other artifacts. Early use of standard color flow equipment led to saturation and blooming, obscuring small vasculature. More recently, low Mechanical Index (MI<0.15) harmonic techniques have been devised to visualize nearly stationary microbubbles in the microcirculation in real time (frame rates>8hz). These methods provide a fleeting view of the larger feeding vessels at the beginning of a bolus before the microcirculation fills obscuring their view ( 5-10 secs). Moderate power techniques (MI 0.2-0.3) are used to visualize the macrovasculature by destroying the bubbles in the microcirculation, but then perfusion is lost. We present a method using a low-mi pulse-inversion acquisition with ultrasound contrast agents, which enables the simultaneous visualization of blood flow in both the macro- and micro-vasculature. A combination of RF and Doppler filters enables the separation of the fundamental and harmonic flow components from tissue and stationary microbubbles. These flow signals are combined with conventional pulse inversion processing to make an image displaying both perfusion and flow. Using Definity at an MI of 0.07, this technique was compared against color flow without contrast for the detection of macrovascular blood flow in a human liver. Scanning at low MI reduced clutter in the fundamental and harmonic signals compared to high MI color flow by 25 db and 45 db respectively. Adding the contrast agent, even at very low MI, increased flow signals relative to color flow for the two components up to 15 db and 4 db, respectively. This gives overall improvement of flow to clutter ratios of 40 db and 49 db for the fundamental and harmonic components respectively, yielding improved flow sensitivity and penetration. 182
1E-2 10:45 a.m. DETECTION OF HARD-SHELLED ULTRASOUND CONTRAST AGENTS USING A SIGNAL SUBTRACTION METHOD S.H.BLOCH*,D.N.PATEL,P.A.DAYTON,M.WAN,andK.W.FERRARA, Department of Biomedical Engineering, UC Davis. Corresponding e-mail: shbloch@ucdavis.edu Hard-shelled contrast agents demonstrate unique properties because the enclosed gas oscillates freely only after the shell is damaged. We have observed this behavior optically using a high-speed camera, and here we exploit this property to improve the ultrasound differentiation of hard-shelled microbubbles and tissue. Previous investigations have shown that microbubbles can be detected using phase inversion techniques. If one-cycle pulses are transmitted, the 180 (rarefaction-first) bubble echo contains harmonics while the 0 (compressionfirst) echo does not. This effect does not depend on the order of pulse transmission in flexible-shelled bubbles. However, we have previously reported that a submicron polymer-shelled agent, M1091, does exhibit order-dependent effects: the first echo is lower in amplitude and higher in frequency than the second echo, regardless of transmitted phase. We have now confirmed these effects using a second polymer-shelled agent, BG1135, and believe these to be general properties of hard-shelled microbubbles, whose response to the first ultrasound pulse includes the disruption of the microbubble shell. Based on these findings, we were able to design a pulse sequence to detect hard-shelled bubbles, signal subtraction, in which two same-phase single-cycle pulses are transmitted in succession and the resulting echoes subtracted. This subtraction results in the cancellation of tissue echoes even in the presence of tissue harmonic generation, with only a small reduction in bubble signal amplitude. Using 2.25 MHz, 1 MPa pulses with 180 phase, we achieved agent-to-tissue ratios exceeding 30 db, compared to less than 20 db for single-cycle phase inversion. We have studied the effectiveness of this technique relative to changes in pressure, transmitted frequency, and bubble and tissue motion. This research was supported by POINT Biomedical Corp., Bracco Research S.A., and NIH grant CA76062. The authors thank Peter Frinking of Bracco Research, and Robert Short and Glenn Tickner of POINT Biomedical for their assistance. 1E-3 11:00 a.m. IN-VITRO B-MODE CONTRAST IMAGING USING CHIRP EXCITATION WITH A NON-LINEAR DECODER J. BORSBOOM*, C. TING CHIN, and N. DE JONG, Erasmus MC, Rotterdam, The Netherlands. Corresponding e-mail: j.borsboom@erasmusmc.nl Coded excitation using chirps is used in ultrasound imaging to increase signal-tonoise ratio (SNR) and penetration depth. Previously, we proposed a non-linear 183
compression filter that can selectively compress and extract the fundamental and second harmonic parts of received echoes and obtain axial resolutions comparable to conventional imaging. In this study, we investigated whether non-linear contrast imaging with chirp excitation improves 2-D image quality due to increased SNR and increased contrast-to-tissue ratio (CTR). To quantitatively evaluate the performance of the non-linear coded excitation technique, we made B-mode images of a flow phantom using pulse and chirp excitations in both fundamental and second harmonic mode and compared the results by eyes, SNR and CTR. The flow phantom consisted of tissue-mimicking material with embedded flow channels of 10 mm, 5 mm, and 1 mm diameters in which either saline or an experimental contrast agent (BR14, Bracco Research SA, Geneva, Switzerland) was flowing. For scanning we used a 3.5 MHz, 65% bandwidth unfocussed transducer in pulse-echo mode excited with 0.98 µs Gaussian pulses and 10 µs Gaussian apodised chirps, both with centre frequency 2 MHz, 45% bandwidth and, MI s up to 0.15. The transducer was mechanically scanned over the surface of the phantom. The echoes were digitized and recorded with a LeCroy digital oscilloscope and processed off-line. The resulting 2-D images showed the 10 and 5 mm channels in both fundamental and second harmonic modes. The spatial resolution of the harmonic chirp image is comparable to the spatial resolution of the other modes. SNR improved by 9.8 db for the fundamental and 8.8 db for the harmonic image when changing from pulse to chirp excitation. These figures compare well to an expected increase of 10 db. The CTR was found to be approximately equal for both pulse and chirp excitation. Hydrophone measurements showed that propagation harmonics generation is mainly dependent on peak pressure. Therefore, chirp excitation might be used to decrease contaminating tissue harmonics and hence increase CTR by decreasing the transmitted MI. 1E-4 11:15 a.m. A NEW IMAGING STRATEGY UTILIZING WIDEBAND TRANSIENT RESPONSE OF ULTRASOUND CONTRAST AGENTS D. E. KRUSE* 1,C.K.YEH 2, and K. W. FERRARA 1, 1 University of California, Davis, CA, 2 National Taiwan University, Taipei, Taiwan. Corresponding e-mail: dekruse@ucdavis.edu High resolution clinical systems operating near 15 MHz are just becoming available; however they lack sensitive harmonic imaging modes for contrast agent detection primarily due to limited bandwidth. When sufficiently driven, it is well known from both experiment and theoretical models that microbubbles produce omnidirectional wideband acoustic transient responses upon collapse. This response has been utilized at frequencies below 7 MHz, but has not been explored at frequencies extending to 15 MHz, where resolution and sensitivity degrade more quickly as a function of depth. To overcome these limitations, we propose a novel strategy utilizing 2 separate transducers at widely separated frequencies 184
and arranged confocally for transmit 2 MHz/receive 15 MHz (T2R15) to simultaneously excite and receive acoustic transients. Experiments were performed to demonstrate that T2R15 shows similar resolution and greatly reduced attenuation compared to T15R15, and superior resolution compared to T2R2. In the experiments, T2 consisted of a 2.25 MHz single cycle, rarefaction-first pulse, with peak negative pressure varied from -200 kpa to -1.2 MPa. For comparison, T15 was generated with a wideband excitation and peak negative pressure of -240 kpa to simulate the pressure obtained following 4 cm of an attenuating medium (0.5 db/cm/mhz). To gain perspective, this pressure corresponds to -7.6 MPa without attenuation and neglecting non-linear propagation. To simulate closely spaced vessels, a lipid shelled contrast agent was pumped through two, 200 µm tubes with centers separated by 400 µm with a mean velocity of 2 mm/s. The flow phantom was placed in a water tank at the focus of both transducers, which were separated by 40 and mechanically scanned together. The axial resolution for T2R15 was sufficient to resolve and map Doppler power flow profiles in both vessels, thus illustrating high resolution and bandwidth of T2R15, which would be impossible at T2R2, where λ=770 µm. Acoustic transients were detected for T2 starting near -400 kpa. Results comparing T2R15 bandwidth and echo amplitude as a function of transmitted pressure will be presented. We acknowledge the NIH (EB 00239) for their support. 1E-5 11:30 a.m. TRIPLET PULSE SEQUENCE FOR SUPERIOR MICROBUBBLE/TISSUE CONTRAST S. UMEMURA* 1,T.AZUMA 1, H. KURIBARA 2, and H. KANDA 2, 1 Hitachi Central Research Laboratory, 2 Hitachi Medical Corporation. Corresponding e-mail: sumemura@crl.hitachi.co.jp A number of imaging sequences have been proposed to improve microbubble/tissue contrast of harmonic imaging, in which the intensityof tissue harmonic echoes, arising from the nonlinear propagation of a transmit pulse in tissue, can be too significant to ignore in comparison with that of harmonic echoes from microbubbles. It is considered that the phase relation between transmit waves and reflected tissue echoes remains constant even after nonlinear propagation in the diagnostic amplitude range. In contrast, microbubble oscillation is so nonlinear that the phase relation between transmit waves and microbubble echoes may vary depending on the transmit amplitude. A new pulse echo sequence is conceived utilizing this difference of nonlinear character for surperior microbubble/tissue contrast. Three-time transmit/receive are performed with transmit pulses with the same envelope and a phase shift by 120, and the received echoes are summed. In the summation, the fundamental and second-harmonic components of tissue echoes will be completely cancelled while those of microbubble echoes will persist. The first order estimation of the microbubble/tissue contrast achievable with the proposed sequence was carried out, based on the numerical simulations of microbubble scattering and nonlinear propagation as responces to 185
the same pulse waves at 2 MHz. Scattering by a microbubble in water subjected to an acoustic pulse was calculated by numerically solving the Rayleigh-Plesset equation. Nonlinear propagation of an acoustic pulse in a contiguous tissue with the same attenuation coeffcient and B/A parameter as human liver was calculated using a finite element code, PZFlex. Waveforms after travelling 60 mm were analyzed. The transmit amplitude was adjusted so that it would be reduced to a certain magnitude at the reference point. Both calculated echo waveforms were purified with the same band pass filter, and the output peak-to-peak amplitudes were compared. The result predicted that the new sequence would achieve a microbubble/tissue contrast superior to the conventional B-mode and the pulse inversion sequence by 70 and 40 db, respectively, at a pulse-peak amplitude of 0.1 MPa. The experimental results with a tissue phantom and those with a flow phantom will also be presented. Authors thank to K. Asafusa and T. Osaka for their cooperation in constructing the data acquisition system for the experiment. 1E-6 11:45 a.m. A NEW HIGH FREQUENCY DESTRUCTION/REPERFUSION SYSTEM C. K. YEH* 1, D. E. KRUSE 2, M. C. LIM 3, D. E. REDLINE 3, and K. W. FERRARA 2, 1 National Taiwan University, Taipei, Taiwan, 2 University of California, Davis, CA, 3 UC Davis Medical Center, Sacramento, CA. Corresponding e-mail: cyeh@ucdavis.edu In order to improve the resolution of contrast-assisted imaging systems, we have created a high frequency destruction/reperfusion imaging system with a spatial resolution of 160 µm x 160 µm. This represents a 25x improvement in resolution compared to previous efforts at clinical frequencies ( 800 µm square), thus allowing a separation of slow capillary and faster arteriole flow velocities. The system utilizes a 1-MHz cylindrically focused transducer for destruction and a 25-MHz spherically focused transducer for pulse/echo imaging. New signal processing methods take advantage of a priori knowledge of contrast echo decorrelation between frames as an alternative to previous methods which use non-linear bubble oscillations to separate bubble and tissue echoes. Speckle tracking and a clutter filter are applied across frames to remove the challenging physiologic motion artifacts that are obtained when imaging with a mechanically scanned transducer. To test the system, a new flow phantom has been developed capable of generating very low flow velocities typically found in the microcirculation ranging from 0.1 to 5 mm/s. Using an exponential reperfusion model, flow constants proportional to absolute flow rate were estimated from B-mode time-intensity curves. The in vitro results indicate a strong correlation (R=0.99) between the actual flow velocity and the estimated rate constant. The time intensity curves for both the in vitro and in vivo data show similar variability over time which is shown to be a function of the region-of-interest (ROI) size over which the intensities are summed. Preliminary in vivo images are presented showing blood perfusion in the ciliary processes and iris of the rabbit 186
eye. We estimate the time to 80% intensity for the in vivo data and demonstrate that this time ranges from 0.8 to 6.7 s. ROIs from within the iris include a range of reperfusion times expected for arterioles and capillaries. ROIs from the ciliary processes yielded slower perfusion as expected from vascular casts of the microcirculation in this region. Potential applications of this system include high-resolution perfusion assessment in small animals. We acknowledge the NIH (CA 76062) for their support. Session: 2E ALL YOU WANTED TO KNOW ABOUT IMAGING Chair: T. Thomas Siemens Ultrasound 2E-1 10:30 a.m. (Invited) CODED EXCITATION FOR DIAGNOSTIC ULTRASOUND: A SYSTEM DEVELOPER S PERSPECTIVE R. Y. CHIAO* and X. HAO, GE Medical Systems, Milwaukee, WI. Corresponding e-mail: Richard.Chiao@med.ge.com Image quality in diagnostic B-mode ultrasound is fundamentally determined by resolution and penetration. Resolution depends on the frequency and bandwidth of the transmitted pulse, while penetration depends on the pulse energy, which in turn depends on the pulse amplitude and duration. Conventionally, high bandwidth for resolution has been achieved by using a short pulse. This results in a tradeoff between resolution and penetration since a short pulse also constrains the pulse energy due to the limited amplitude that may be transmitted for regulatory or hardware reasons. Coded excitation increases the transmitted pulse duration for increased energy while retaining the resolution of a short pulse through appropriate coding on transmit and decoding on receive. Physically, pluse energy is distributed with coding over a long time interval and over a wide bandwidth on transmit that is subsequently compressed to a short time interval on receive to gain SNR without loss of resolution. Coded Excitation has long been used in RADAR, however it has only recently been implemented on commercial ultrasound systems. This paper reviews the basic concepts behind coding and decoding for pulse compression including binary and chirp coding for both fundamental and harmonic imaging; discusses practical implementation issues related to tranducer bandwidth, dynamic focusing, and frame rate; and presents clinical images and a new theoretical result that achieves complete 2nd harmonic compression simultaneous with fundamental cancellation by using quadrature signals. 187