On the relationship between encapsulated ultrasound contrast agent and pressure

On the relationship between encapsulated ultrasound contrast agent and pressure

Ultrasound in Med. & Biol., Vol. 31, No. 5, pp. 673– 686, 2005 Copyright © 2005 World Federation for Ultrasound in Medicine & Biology Printed in the U...

663KB Sizes 4 Downloads 38 Views

Ultrasound in Med. & Biol., Vol. 31, No. 5, pp. 673– 686, 2005 Copyright © 2005 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/05/$–see front matter

doi:10.1016/j.ultrasmedbio.2005.01.005

● Original Contribution ON THE RELATIONSHIP BETWEEN ENCAPSULATED ULTRASOUND CONTRAST AGENT AND PRESSURE DAN ADAM, MICHAL SAPUNAR and ELINA BURLA Department of Biomedical Engineering, Technion – I.I.T., Haifa, Israel (Received 29 July 2003, revised 29 December 2004, accepted 6 January 2005)

Abstract—Noninvasive measurement of pressure within the heart cavities and other internal organs (e.g., kidney, liver) has significant clinical value, but currently is not feasible. Noninvasive pressure estimation using encapsulated ultrasound (US) contrast agents (UCA) as sensors is a challenge because they supposedly respond to their ambient pressure, but they are more rigid and less sensitive to pressure than gas microbubbles. Here, Optison® sensitivity was studied (fresonance ⴝ ⬇ 2 MHz) to varying pressures, when excited at 2 times and also at 0.5 times fresonance. Cyclic momentary increases in ambient pressure of 0 to 5, 0 to 10, 0 to 15 or 0 to 20 kPa at 1.0 Hz, mimicking left ventricular (LV) pressure changes, caused amplitude decrease of echoes at 0.5, 1 and 2 times the transmitted frequency and decrease of attenuation. Changes at 0.5 times the transmitted frequency correlated best, but only after 70 to 150 s. The correlations (mean ⴞ SD) during 150 to 300 s were 0.706 ⴞ 0.072 for 0 to 10 kPa, 0.844 ⴞ 0.042 for 0 to 15 kPa and 0.859 ⴞ 0.031 for 0 to 20 kPa. Attenuation presented less correlation. For 1.0 Hz, 10 to 15 kPa or 15 to 20 kPa pressures, mimicking systemic pressures, the attenuation decayed fast and even faster for slow (0.05 Hz) cyclic varying pressures, or elevated steady-state pressures (of 10 kPa and 20 kPa). Thus, cyclic pressure effects on UCA are demonstrated to be reversible, but elevated static pressures cause UCA destruction. This allows cyclic pressure variations to be detected, using the subharmonics of the transmitted frequency, down to 10 kPa. (E-mail: [email protected]) © 2005 World Federation for Ultrasound in Medicine & Biology. Key Words: Ultrasound, Contrast agent, Pressure estimation, Cardiovascular.

the scattering of US from bubbles and following changes in the scattering properties as the ambient pressure changes (e.g., blood pressure changes) was presented over 20 y ago (Fairbank and Scully 1977; Hok 1981). Yet, the free bubbles used at that time had a wide dispersion of sizes that did not allow measuring their resonance frequency, as suggested by Fairbank and Scully (1977), and the fact that the gas bubbles dissolved within seconds of being injected IV did not allow any practical implementation (Hok 1981; Shankar et al. 1986). Any practical implementation of measuring blood pressure in the body, based on microbubbles, should include a means of providing the microbubbles to the region-of-interest (ROI) (e.g., Bouakaz et al. 1999) and development of microbubbles with such properties that will allow stationary or quasistationary conditions for at least one or several heart cycles. Production of microbubbles by cavitation, for example (Miwa 1984), is dangerous and does not allow stable measuring conditions because their dissolution is too fast. During the last decade, stable encapsulated microbubbles have been developed for the purpose of functioning as US contrast

INTRODUCTION Currently, pressure measurements within the heart cavities and other internal organs (e.g., kidney and liver) are mainly performed invasively by a pressure catheter, or noninvasively by means of Doppler ultrasound (US) using the simplified Bernoulli equation. The invasive methods provide trustworthy results, but are accompanied by pain and risk of infection. The noninvasive approach, unfortunately, does not provide reliable or reproducible blood pressure values (Strauss et al. 1993; Reddy et al. 2003). The noninvasive measurement is usually performed at a convenient location (e.g., arm or finger tip), but the blood pressure values of interest are at other locations (e.g., internal organs, cardiac cavities, etc.). There have been many attempts to develop other alternative indirect noninvasive methods for blood pressure monitoring. The idea of measuring noninvasively

Address correspondence to: Prof. Dan Adam, Department of Biomedical Engineering, Technion – Israel Institute of Technology, Haifa 32000 Israel. E-mail: [email protected] 673

674

Ultrasound in Medicine and Biology

agents (UCAs). The availability of such stable gas microbubbles generated new interest in developing a measuring technique based on the interaction of US with microbubbles injected into or flowing through the region where the value of ambient pressure needs to be measured (Bouakaz et al. 1999; Frinking et al. 2001; Kirkhorn et al. 2001). UCAs have been developed and approved for clinical use for improving the imaging of the vascular system (and, hopefully, also the perfusion of tissue). UCAs consist of gas microbubbles, usually with shells, of diameters of less than 10 ␮m to allow for passage through the lung capillaries. The large difference of acoustic impedance between the gas bubble and that of the surrounding medium produces large backscattered US signals, thus enhancing echoes from the blood in which the microbubbles flow. At the same time, the compressibility of the microbubbles is very different from that of blood. This compressibility allows microbubbles to change considerably in size in response to changes in pressure and, thus, even to demonstrate cyclic changes in videodensity (Shapiro et al. 1990). Variation in the size, in turn, should affect the acoustic characteristics of microbubbles, such as their resonance frequency (Fairbank and Scully 1977), reflectivity or attenuation of the US waves. Therefore, the value of surrounding pressure could be derived from the acoustic characteristics of UCAs by injecting gas-containing bubbles of nearly uniform and known size into the organ in which the blood pressure is to be measured. There are mainly two possible mechanisms of stabilization of the microbubbles, encapsulation with an elastic shell and substitution of the microbubble core by inert gases (Van Liew and Raychaudhuri 1997). Unlike the first-generation UCAs, which were composed of free air bubbles, the second-generation agents implement both methods and include gases of a higher molecular weight, less solubility and diffusivity, encapsulated into organic or polymeric shells. Both of the stabilization methods mentioned above greatly affect the response of the contrast microbubbles when exposed to changes in pressure or to US insonation. It should be emphasized here that, even when stabilized microbubbles are used, there are time-dependent processes taking place from the moment the UCA is injected (e.g., into saline or blood). The response of air-containing encapsulated microbubbles to ambient pressure changes was intensively studied. The effects of pressure on the size changes of Albunex® (Molecular Biosystems Inc., San Diego, CA, USA) have been reported (de Jong et al. 1993), and Quantison™ (Andaris Ltd., Nottingham, UK) was reported to be insensitive to ambient pressure changes due to its rigid shell (Bouakaz et al. 1998). Shi et al. (1999a) demonstrated the nonlinear response of Levovist®

Volume 31, Number 5, 2005

(Schering AG, Berlin, Germany) to the changes in pressure. However, very few attempts were made to evaluate the pressure sensitivity of UCAs with gas cores different from air; Podell et al. 1999a reported on several factors, among them the ambient pressure, that modify the behavior of Optison® microbubbles. When Optison®, a second-generation UCA, is employed, it is important to identify the different diffusion processes between the octafluoropropane (C3F8) gas in the microbubbles and the gases in the ambient liquid. When air bubbles are immersed in liquid (e.g., water or blood, exposed to air), diffusion processes start usually mostly outward, with the rate of diffusion being strongly affected by the partial pressures of the different gases in air and in the liquid, the microbubble surface tension, the systemic blood pressure and the initial mole fraction of the osmotic agent in the bubble (Kabalnov et al. 1998a, 1998b). When microbubbles containing slowly-permeating gas are used, the process is much more complicated. The permeation coefficient of any inert gas, such as the ones used in second-generation UCAs, is significantly lower than the permeation coefficient of the gas components contained in air (Kabalnov et al. 1998a, 1998b). Because the partial pressures of the air components (N2, O2 and CO2) inside the UCA microbubbles are very low, there are large initial partial pressure differences for these gases, which cause inward diffusion of air components. This process is very fast and is completed before significant loss of the inert gas. The partial pressure difference, which causes the outward diffusion of the inert gas, is equal to partial pressure of this gas inside the bubble because, initially, there is no inert gas in the ambient liquid (e.g., water or blood) and the outside partial pressure is assumed to be zero. The initial partial pressure of inert gas inside the microbubbles is actually above the atmospheric pressure (1 atm) because of the additional pressure caused by the surface tension. The entry of air gases dilutes the partial pressure of the inert gas, decreasing the driving force for outward diffusion, but increasing the volume of the microbubble. The growth of the microbubble causes an increase of surface area, which tends to increase the outward diffusion of the inert gas, depending on the mole fraction of the osmotic agent (Kabalnov et al. 1998a, 1998b). The increased volume also causes less curvature of the surface of the microbubble, which tends to decrease the diffusion of inert gas. The net result of these conflicting effects is a decrease of the rate of outflow of the inert gas, to approximately one third of the hypothetical rate if air gases had not entered the microbubbles, thereby prolonging their time of existence (Van Liew and Burkard 1995; Kabalnov et al. 1998a, 1998b). There are various reports that demonstrate that Optison® microbubbles may last almost 20 min. After the bubble reaches the maximal

Contrast agent and pressure ● D. ADAM et al.

size, it slowly shrinks because of the loss of the inert gas; the diffusion of the air gases also exists, of course, but the rate is far slower than it would be if the inert gas was not present to lower their concentrations. While the bubble is being absorbed, the constituent gases tend to stay in constant proportion to each other. Thus, it is evident that the diffusion processes described above dramatically affect the acoustic behavior of the bubbles containing “slow-permeating” gases. The role of any change in the ambient pressure is significant because it changes the balance of the partial pressures of the gases in the surrounding liquid. Although the properties of microbubbles and, specifically, those of UCAs, can be studied using optical techniques (de Jong et al. 2000), these techniques are usually applied to small quantities under a microscope and cannot be used within tissue, where US techniques are preferable. Ultrasonic energy, when used at high levels, could produce a “rectified-diffusion” effect, burst the microbubbles and even cause tissue damage through this mechanism (Hynynen et al. 2003). When US is used at low energy levels sufficient for imaging (e.g., mechanical index, MI, ⬍ 0.2, the ratio of the maximal US rarefactional pressure in MPa to the square root of the transducer frequency in MHz), and also for short durations, it can be used as a measurement tool with little influence on the measured quantity (UCAs). In order better to understand the behavior of the UCA microbubbles under different ambient pressures, in this study, three conditions have been studied: ● ● ●

UCA behavior under “fast” (physiological) cyclic changes of pressure UCA behavior under “slow” cyclic changes of pressure UCA behavior under different constant elevated pressures.

Although the last part of the problem has been intensively investigated and some promising results have been obtained (Shankar et al. 1986; Shi et al. 1999b), only a few attempts have been made to study the problem of estimating the dynamic changes. Because the natural frequencies of the microbubbles are much higher than the frequency of the pressure fluctuations, the behavior of the microbubbles can be considered to be “quasisteady” (Ran and Katz 1991). This assumption may be hardly accepted for air-contained UCAs. An accurate model of the acoustic response of microbubbles containing inert gas should incorporate, therefore, the US insonation and the diffusion processes that are affected by cyclic changes in ambient pressure. Some studies were concerned with the problem of time-varying pressure (Ran and Katz 1991; Padial et al. 1995; Brayman et al. 1996; Greim et al. 2000) and its effect on the intensity of

675

the US image or its brightness. Low correlation was found. The present study was undertaken to determine the effects and to understand the mechanisms of applied pressure on the acoustic scattering and attenuation of Optison® suspensions by means of in vitro experiments. Measuring the attenuation parameter will not be the optimal measurement for blood pressure estimation in the clinical setting, because two separate transducers are required, one for US insonation and the second, for reception. These measurements, however, are important for understanding the processes involved in interactions between the UCA, the liquid, the US pulses and the ambient pressure. METHODS The UCA Optison® (at the time of study, Mallinckrodt Medical GmbH, Hennef, Germany) was used in the current experiments. Optison® has been approved (by the US Food and Drug Administration, FDA, by the European Agency for Evaluation of Medicinal Products, EMEA) for use in clinical echocardiography. Optison® consists of a sterile suspension of human serum-albumincoated microspheres filled with octafluoropropane (C3F8) with a concentration of 5.0 to 8.0 ⫻ 108 bubbles/mL. The microbubbles have a mean diameter in the range of 2.0 to 4.5 ␮m, with 93% of the microbubbles being smaller than 10 ␮m. The approximate amount of octafluoropropane gas in each mL of Optison® is 0.22 mg. Optison® was injected into saline solution using a 20 G syringe, at a maximum injection rate of less than 1.0 mL/s. To limit excessive acoustic attenuation and multiple scattering, a diluted solution of Optison® in saline (0.9 % NaCl) with a concentration of 0.1 ␮L/mL was used. A slowly rotating magnetic stirrer kept it in circulation. All experiments were performed in buffer solution at 22 ⫾ 2 °C. The resonance frequency of Optison® has been measured to be around 1.9 to 2.0 MHz (Chen et al. 2002) (see below). Efficient responses from the microbubbles can be achieved when the excitation is done at the resonance frequency, at half or at double the resonance frequency (Shi and Forsberg 2000). Excitation at the resonance frequency is efficient, but causes rapid rupture of the microbubbles and is not recommended (Chomas et al. 2002). Excitation at double the resonant frequency is preferable for detection of UCAs in tissue (because tissue does not produce half harmonics in the way that a UCA does) (Shankar et al. 1999; Shi et al. 1999a) and, therefore, the interrogation was done with US pulses of nine cycles in duration and at a frequency of 4 MHz. Because the UCA was immersed in water with no tissue around, in some experiments, the excitation was done at

676

Ultrasound in Medicine and Biology

Volume 31, Number 5, 2005

capable of measuring both ultrasonic attenuation and scattering. It consisted of: ●

Fig. 1. The in vitro experimental set-up block diagram and pressure container.

half the resonance frequency (i.e., at a frequency of 1 MHz) and with US pulses of 16 ␮s in duration. The experimental set-up The block diagram of the experimental set-up is illustrated in Fig. 1. The pressure container was made of Perspex and its size was 17 cm high ⫻ 13 cm long ⫻ 13 cm wide. The container was filled with 0.5 L saline solution. The container consisted also of an inner 12-cm tall rectangular “chamber,” approximately 4 cm by 4 cm in size, made of thin polyethylene. This inner chamber, which was placed about 4 cm away from the surface of the US transducer/receiver, was also filled with the saline solution, and the UCA suspension was injected only to this chamber and was slowly stirred by a magnetic stirrer. The pressure container included two holes for placing the US transducers/receivers, with proper sealing. The inlet and outlet of the container were constructed so as to allow injection of UCA microbubble suspensions into the inner chamber and for applying and controlling the external pressures. The pressure inside the container was monitored by a digital pressure switch ISE4LB01-65 (SMC Corporation of America, Indianapolis, IN, USA) pressure meter with accuracy of 1 kPa (7.5 mmHg) and adjusted by precision regulator IR1000-01B-R (SMC Pneumatics). Varying air pressure above the surface of the liquid changed the pressure in the saline. Pressure gauge (SMC, ZSE40F) allowed feedback control of the pressure changes, run by a LabView® program. A pulse-echo measurement system was assembled,







The RITEC advanced measurement system RAM 5000, which includes a high-power gated radio-frequency (RF) amplifier, an internal pulse generator and a broadband receiver (“RITEC”). Usually, US pulses of nine cycles in length were used, at 4 MHz, 200 kPa, with PRF ⫽ 10 Hz. In some experiments, the pulses were of 16 cycles in length, at 1 MHz, 50 kPa, with PRF ⫽ 5 Hz. The transmitted signals were first amplified and then supplied to the acoustic transducer. The amplitude of the pulse was adjusted by the internal amplifier gain and a high-power attenuator RITEC RA-30, to produce the acoustic pressures as measured 5 cm away from the surface of the transducer, beyond the near-field distance, at the location where the 4 cm by 4 cm inner chamber was later placed. The pressure measurements were made by a hydrophone (see below). The pressures used during the experiments were made low enough so as to decrease the possibility of microbubble destruction (Chen and Shung 1998) (e.g., for the 1-MHz signal, an acoustic pressure of ⬇ 50 kPa or MI ⬇ 0.05). An acoustic single-element focused transmitter/receiver transducer, V310 (Panametrics, Waltham, MA, USA), with a central frequency f0 ⫽ 5.5 MHz and 111.6% band width and near-field distance of 3.3 cm. For those experiments in which a 1-MHz signal was transmitted, a focused transducer was used, V303 (f0 ⫽ 0.95 MHz and 61.54% band width). Either of the transducers was placed in a hole in the container wall (transducer). Hydrophone model: PVDF-GL-0400IA (Specialty Engineering Associates, Soquel, CA, USA), supplied with a 17-dB amplifier, was placed in the hole in the container wall precisely opposite the transducer (hydrophone) for the attenuation measurements. An A/D sampling system, which consisted of the CompuScope CS14100 (Gage Applied Technologies, Inc., Lachine, QC, Canada), characterized by 14-bit resolution and 50-MHz sampling rate, connected to a Pentium III personal computer (PC) (digitizer). The A/D system was triggered by the transmitted signal and the response was measured every 0.1 s for a 130.5-␮s duration of data.

Processing of the signals The measured signals were off-line processed and their power spectra were estimated. The acquired backscattered echoes were measured beyond the near-field distance of the transducer and within the 4 cm by 4 cm chamber filled with Optison®, so as to minimize the effects of attenuation (and changes of attenuation during

Contrast agent and pressure ● D. ADAM et al.

the experiment) along the path of the US waves. Total power of the passed-through signals allowing calculation of the attenuation, was defined as: Attenuation[dB]



⫽ 10 · log10

兺 (Signal[mV]reference)2 ⁄ length(signalreference) 兺 (Signal[mV]Optison)2 ⁄ length(signalOptison)



where the path-length is calculated throughout the inner chamber (i.e., 4 cm). Within the estimated power spectrum, the amplitudes of echoes at x0.5, x1, x2 of the imposed frequency were measured and presented as a function of time vs. the ambient pressure changes. The amplitude at each of these frequencies (e.g., for transmission at 4 MHz, the echoes were measured at x0.5 [i.e., at “half” harmonic ⫽ 2 MHz], at x1 [i.e., at “first” harmonic ⫽ 4 MHz] and at x2 [i.e., at the “second” harmonic ⫽ 8 MHz]) was measured by searching for the peak amplitude within ⫾ 100-kHz band around the specified frequency. In order further to reduce the effects of attenuation on the measured amplitudes of the echoes, specifically the one of interest at x0.5 of the imposed frequency, the ratio of amplitudes at x1 harmonic to those at x0.5 harmonic was calculated as a function of time. Also, the correlation of those ratios of echo signals to the pressure changes was calculated (Fig. 5). The experimental procedure The pressure container was filled by 0.5 L of saline at room temperature (22 °C), Optison® was slowly injected and the container was closed. At 30 s after injection, so as to ensure proper UCA mixing, a LabView® program was started that controlled the pressure sequence and, concurrently, the US measurements. This corresponds to time ⫽ 0 in the results below. The measurements were carried out for 5-min duration and performed every 1 s. (In some experiments, those of slow pressure changes and those of constant pressures, the pressures were imposed for 10 min and the US measurements were made every 2 s.) Each experiment was repeated nine times and the averaged results are reported. After each experiment at a given set of parameters, the container was drained, flushed and fresh saline and Optison® were used. Calibration measurements were made by filling the chamber with saline and by measuring the transmitted pulses using the hydrophone. “Fast” (physiological) cyclic changes of pressure. Experiments were performed in which air-saturated saline samples containing UCAs were subjected to timevarying ambient pressures, Patm with air pressure changes of 0 to 5 kPa, 0 to 10 kPa, 0 to 15 kPa or 0 to 20 kPa, imposed at 1.0 Hz, mimicking left ventricular

677

(LV) pressure changes. Also, experiments were performed where the pressure changes were 10 to 15 kPa and 15 to 20 kPa, mimicking aortic pressure changes (where 20 kPa ⫽ 150 mmHg). “Slow” cyclic changes of pressure. In order better to understand the influence of the pressure changes on the diffusion processes and the acoustic behavior of microbubbles, the elevated pressure was also cyclically modulated at slower frequencies of 0.05 Hz, with amplitudes of 0 to 10 kPa and 0 to 20 kPa. The transmitted frequency, in this case, was 1 MHz, MI ⫽ 0.2 and PRF ⫽ 10 Hz. UCA behavior under different constant elevated pressures. Measurements were also made at different values of static pressure: P ⫽ Patm ⫹ Pi, where Pi is either 0, 5, 10, 15 or 20 kPa. These sequences of experiments were performed by increasing the pressure for 20 min, while measuring the echoes to the transmitted signals, as well as the acoustic attenuation when the Optison® was injected, diluted to 0.1 ␮L/mL, into air-saturated saline. Oxygen concentration was 6 ⫾ 1 mg/L and the saline temperature was 22 ⫾ 1 °C. The measurements were repeated when the Optison® was injected into degassed saline, degassed for 24 h at ⫺0.6 bar (oxygen level of 3.5 ⫾ 0.5 mg/L); although such water still contains gases, the effects of these low concentrations are already very significant (see below). The measurements were also repeated when the degassed saline was perfused for 12 h with either nitrogen (PN2 ⬎ 600) or pure oxygen (8.8 ⫾ 0.6 mg/L). RESULTS UCA behavior under “fast” (physiological) cyclic changes of pressure A typical echo response of the Optison® diluted in saline to 0 to 20 kPa cyclic pressure changes can be seen in Fig. 2. The Optison® is excited by US pulses at 4 MHz, 200 kPa. In this figure, the measured pressure changes (plotted not in scale) are depicted, together with the amplitudes of the echoes at either (a) x1 (“first harmonic”), (b) x2 (“second harmonic”) or (c) x0.5 (“half harmonic”) of the frequency of the transmitted US pulses. In each of these panels, both a condensed graph and a zoomed-in section of 20 s are depicted. The amplitudes, at the specific frequencies, are specified in dB, with respect to their integral over the whole frequency range, when measured from the liquid before Optison® was added. The amplitudes of the echoes at x1 and x2 of the transmitted pulses seem to be quite noisy, with little correlation to the imposed cyclic pressure changes, as also verified by calculating the correlation coefficient series between each two (see Fig. 3). The amplitude of the echoes at x0.5 the transmitted frequency is highly

678

Ultrasound in Medicine and Biology

Fig. 2. The first, second and subharmonic response (left ordinate: amplitude [dB]) of Optison® diluted in saline under cyclic pressure of 0 to 20 kPa (right ordinate: pressure [kPa]), as a function of time (s). Transmitted frequency 4 MHz, 200 kPa (MI ⫽ 0.1). (Left) Data from the whole experiment; (right) zoom on 20 s (100 to 120 s). In each graph, the measured ambient pressure and the amplitude of the echoes are plotted, at the (a) first harmonic, SD ⫽ ⫾ 4.2 dB, (b) second harmonic, SD ⫽ ⫾ 3.9 dB and (c) subharmonic frequencies, SD ⫽ ⫾ 3.6 dB.

correlated (⬎ 0.7) to the pressure changes; see Table 1, although it takes about 70 s for this correlation to build up (Fig. 3). There is also a marked difference in the SD of the mean correlation coefficient series, between the results measured at x0.5, which are low, to those measured at either the x1 or x2 harmonics of the transmitted frequency (bars in Fig. 3). These SD demonstrate also the SD of the measured amplitudes (as depicted in Fig. 2), because the pressure waves were imposed by a pump and had almost no deviation from their mean. The high correlation between the amplitude of the echoes at x0.5 the transmitted frequency to the pressure changes during a certain time period (e.g., 200 to 300 s after Optison® injection) was also obtained when the imposed cyclic pressure changes were in the range of 0 to 15 kPa, 0 to 10 kPa and 0 to 5 kPa. The results of these experiments are summarized in Table 2. The attenuation of the US pressure field while passing through the 4-cm inner chamber, as defined earlier, was also measured as a function of time, when the imposed cyclic pressure changes were in the range of 0

Volume 31, Number 5, 2005

to 20 kPa, 0 to 15 kPa, 0 to 10 kPa and 0 to 5 kPa. The results of these experiments (the first and last ones) are depicted in Fig. 4. Because the attenuation is altered by the imposed cyclic changes, it also affects the amplitude of the echoes because the echoes are obtained not only from the surface of the chamber containing the Optison®, but also from the more central parts of the Optison® cloud. In order to reduce these effects, the ratios of amplitudes at x1 harmonic to those at x0.5 harmonic were calculated as a function of time (Fig. 5). The averaged correlation coefficients (⫾ SD) of the amplitude ratios (Fig. 5) are also calculated and compared with the averaged correlation coefficients of the amplitudes, as a measure of these cross-effects. A different set of experiments was carried out in which the pressure changes mimicked the systemic blood pressure; changes were made between 10 and 15 kPa and between 15 and 20 kPa. The behavior of the echoes at x0.5 harmonic of the transmitted frequency is depicted in Fig. 6a and b, for both pressure changes. The averaged correlation coefficients between the x0.5 harmonic amplitude and the imposed cyclic pressure are given in Table 2. The behavior of the attenuation as a response to pressure changes made between 10 and 15 kPa is plotted in Fig. 6c and to pressure changes between 15 and 20 kPa is plotted in Fig. 6d. The increase in pressure during the cyclic pressure changes caused not only increase of the amplitude at x0.5 harmonic of the transmitted frequency, but also a shift in the frequency of peak amplitude around the x0.5 harmonic frequency. The averaged frequencies of these local peak amplitudes, within the neighborhood of the

Fig. 3. Averaged correlation coefficients (and their SD) between the harmonic amplitude of echoes from of Optison® diluted in saline and the cyclic pressure, as a function of time. Transmitted frequency 4 MHz, 200 kPa (MI ⫽ 0.1). (a) First harmonic, (b) second harmonic and (c) subharmonic.

Contrast agent and pressure ● D. ADAM et al.

679

Table 1. Average correlation coefficient (ACC) and its SD (of ACC) between the different harmonic amplitudes measured from Optison® diluted in saline and the cyclic pressure (within the time interval of 100 –300 s)

0–20 kPa cyclic pressure changes @ 100–300 s

ACC SD of ACC

x0.5 harmonic, as measured at the maximal and minimal ambient pressures, are summarized in Table 3 for the different experiments (i.e., for different ranges of pressure changes). UCA behavior under “slow” cyclic changes of pressure In order better to understand the results obtained during the experiments with the 1.0-Hz cyclic changes of pressure and, specifically, the different behavior noted between those experiments in which the pressure was varied from 0 kPa to some higher value (e.g., 5, 10, 15 or 20 kPa) and those experiments in which the pressure was raised from some nonzero value to a higher value (e.g., 10 to 15 kPa or 15 to 20 kPa), some experiments were performed with slower cyclic changes and also at steadystate elevated pressures (see below). Figure 7 demonstrates the behavior of the attenuation parameter when “slow” (0.05 Hz) cyclic changes of pressure (0 to 20 kPa) are applied. The results of a typical example are depicted; the 6-min experiment is described in Fig. 7a, and a section is plotted in Fig. 7b. UCA behavior under different constant pressures The echoes of the transmitted 4-MHz US pulses obtained from Optison® diluted in saline were measured as a function of time, at several values of constant ambient pressure. The averaged results of the amplitudes (and their SD in some instances) of the echoes, measured at x0.5 harmonic, x1 harmonic and x2 harmonic of the transmitted frequency, are depicted in Fig. 8a– d. The

⫻1 harmonic

⫻2 harmonic

⫻0.5 harmonic

⫺0.3 0.08

⫺0.027 0.07

⫺0.76 0.05

amplitudes, at the specific frequencies, are specified in dB, with respect to their integral over the whole frequency range, when measured from the liquid before Optison® was added. Although, at a pressure of 5 kPa (Fig. 8a), the echoes at all measured frequencies remain nearly at the same level during the whole experiment (20 min here); at higher pressures, the echoes deteriorate with time. Already, at an elevated pressure of 10 kPa (Fig. 8b), the x1 harmonic and x2 harmonic have deteriorated by nearly 20 dB after some 16 to 17 min. The echoes at the subharmonic frequency demonstrate a more complex behavior, which includes an initial fast deterioration, then some increase of the amplitude that peaks earlier at increased pressures (at about 6 min for a pressure of 10 kPa, 4 min for 15 kPa and 3 min for 20 kPa). When the attenuation is studied during these experiments, it is noted that, at normal pressure, the attenuation deteriorates very little and nearly linearly during the experiment (see below) and even an increase of the pressure by 5 kPa has only a minor effect; the deterioration is all together about 1.2 dB/cm (Fig. 9). A pressure increase of 10 kPa above normal pressure, though, has already a marked effect on the attenuation; after some slow reduction, the attenuation increases and peaks at about 2 to 3 min and then drops fast to nearly zero by 10 to 12 min. At a pressure increase of 15 kPa, the peak is

Table 2. Averaged correlation coefficients (ACC) and its SD (of ACC) between the subharmonic amplitude and the cyclic pressure changes, for different levels of pressure changes imposed on of Optison® diluted in saline (within the time interval indicated)

Cyclic pressure changes

ACC of ⫻ 0.5 harmonic

SD of ACC of ⫻ 0.5 harmonic

0–20 kPa @ 100–300 s 0–15 kPa @ 200–300 s 0–10 kPa @ 200–300 s 0–5 kPa @ 200–300 s 10–15 kPa @ 100–150 s 15–20 kPa @ 100–150 s

⫺0.76 ⫺0.77 ⫺0.68 ⫺0.4 ⫺0.63 ⫺0.5

0.05 0.05 0.1 0.1 0.09 0.09

Fig. 4. The attenuation behavior (in dB/cm) of Optison® diluted in saline under different cyclic pressures, as a function of time. Transmitted frequency 4 MHz, 200 kPa (MI ⫽ 0.1). (Left) Complete experiment; (right) zoom on 20 s (220 to 240 s). (a) 0 to 20 kPa, (b) 0 to 5 kPa.

680

Ultrasound in Medicine and Biology

Volume 31, Number 5, 2005

DISCUSSION

Fig. 5. (a) The ratio of amplitudes at x1 harmonic to that at x0.5 harmonic of Optison® diluted in saline as a function of time, depicted together with the imposed ambient pressure changes (right ordinate). Transmitted frequency 4 MHz, 200 kPa (MI ⫽ 0.1). Zoom on 20 s (100 to 120 s) from the whole experiment. (b) Averaged correlation coefficients (⫾ SD) of the amplitude ratio of (a) as a function of time.

earlier (1.5 to 2 min) and the reduction in attenuation is just as fast. At 20 kPa, the peak is more pronounced and is earlier (about 1 min), followed by a fast reduction (see Fig. 9). The SDs in these experiments were quite high, due to the sensitivity to the exact initial concentration of the Optison®. Because diffusion processes have been cited as causing changes in size of the microbubbles, thus affecting both the echoes and the attenuation, a set of experiments was conducted in which the gas content of the saline was modified. In these experiments, the transmitting frequency was 1.8 MHz, very near to the resonance frequency of the Optison®, the acoustic pressure was 130 kPa (MI ⫽ 0.1) and the rate was PRF ⫽ 10 Hz. The results of these experiments can be seen in Fig. 10. The attenuation of Optison® diluted in saline, which was first degassed and then saturated with nitrogen (after perfusion of 12 h with nitrogen, PN2 ⬎ 600), is plotted as the function of time in Fig. 10a for atmospheric and elevated (20 kPa) static pressures. The experiment was repeated, but with the saline saturated with oxygen (after perfusion of 12 h with oxygen, 8.8 ⫾ 0.6 mg/L), Fig. 10b. The same experiment was repeated, but with air-saturated saline, plotted in Fig. 10c. When degassed saline was used (with oxygen level of 3.5 ⫾ 0.5 mg/L) and all other conditions were the same, the results were very different, as can be seen in Fig. 10d. Because, in this set of experiments, the conditions were somewhat different (1.8 MHz and MI ⫽ 0.1), an experiment was conducted under conditions similar to those of the other experiments reported here (such as in Fig. 10c), so as to allow better comparison; the transmitted frequency here was 4 MHz, MI ⫽ 0.05 and the saline was saturated with air (see Fig. 10e). It is noted that, at normal pressure, the attenuation deteriorates very little and nearly linearly during the 20-min experiment (altogether, by about 0.4 dB/cm; see the graph for P ⫽ Patm in Fig. 10e).

The design of UCAs usually includes a solid shell (e.g., albumin in the case of Optison®,) to provide constant and narrow size distribution of the microbubbles and to stabilize the encapsulated gas, possibly by preventing direct contact between the gas in the microbubble and the surrounding liquid. It has been reported that gases from the Optison® microbubble core are in dynamic equilibrium with the surrounding environment and that the albumin shell does not restrict the diffusion of the gases (Podell et al. 1999a). It has also been shown that, if the shell is dissolved chemically, the encapsulated gases dissipate rapidly; but this is not a situation that exists in vivo, or was studied here in vitro. Also, relatively severe conditions over long periods of time are required to dissolve the albumin shells of the microbubbles (Podell et al. 1999b). On the other hand, the in

Fig. 6. The response of Optison® diluted in saline under cyclic pressure as a function of time. Transmitted frequency 4 MHz, 200 kPa (MI ⫽ 0.1). (Left) Complete experiment; (right) zoom on 20 s. The ambient pressure changes are superimposed on the same graphs, but are not in scale (right ordinate [kPa]). (a) The amplitude of the subharmonic response to changes of 10 to 15 kPa and (b) to changes of 15 to 20 kPa. (c) The attenuation (in dB/cm) as a function of time for 10 to 15 kPa changes and (d) for 15 to 20 kPa.

Contrast agent and pressure ● D. ADAM et al.

681

Table 3. The shift in frequency of the ⫻0.5 harmonic, measured from of Optison® diluted in saline, due to the cyclic increased pressure; averaged frequencies of local peak amplitude in the neighborhood of the ⫻0.5 harmonic Cyclic pressure changes (kPa)

Calculated during (s)

0–20

100–200

0–15

200–300

0–10

200–300

0–5

200–300

10–15

200–300

15–20

100–300

Average (MHz) SD (*105) Average (MHz) SD (*105) Average (MHz) SD (*105) Average (MHz) SD (*105) Average (MHz) SD (*105) Average (MHz) SD (*105)

vivo US signals from Optison® have been reported by many to diminish after some minutes. The results of the present in vitro study demonstrate that, under some conditions, the deterioration is quite fast but, under other conditions, the signals may last for many minutes. The experimental set-up was constructed and utilized for evaluating the echoes and the attenuation parameter as a function of varying pressures added to the hydrostatic pressure. Because attenuation and changes of attenuation affect the values of the measured echoes, these effects should be minimized or corrected for. The container used in the experiments here included an inner tall chamber, 4 cm by 4 cm in size. The container was filled with water and the UCA was injected only into the inner chamber. Thus, the path of the US waves traversed

Fig. 7. The attenuation response (left ordinate, in dB/cm) of Optison® diluted in saline under “slow” (0.05 Hz) cyclic pressure (of 0 to 20 kPa, right ordinate) as a function of time. Transmitted frequency 1 MHz (MI ⫽ 0.2). (a) The complete experiment; (b) zoom on 50- to 150-s period.

Frequency @ max. pressure

Frequency @ min. pressure

2.55 3.5 2.36 3.8 2.18 3.9 1.93 1.74 2.4 4 2.07 3.5

1.9 1.9 1.9 1.75 1.9 1.4 1.92 1.4 2.11 4.2 2.01 2.9

mostly through water. Echoes were measured from regions within the path of the US waves through the UCA mixture in the inner chamber and attenuation was calculated along the whole path (4 cm). The change of attenuation (throughout the whole path) during the whole experiment was usually less than 5 dB. In order to test if the effects of the cyclic pressure changes on attenuation have a significant effect on the echo signals and, specifically, on the correlation of those echo signals to the pressure changes, the ratio of amplitudes at x1 harmonic to those at x0.5 harmonic were calculated as a function of time (Fig. 5). The averaged correlation coefficients (⫾ SD) of the amplitude ratio (Fig. 5) are found to be very similar to those of the x0.5 harmonic; thus, it is con-

Fig. 8. The averaged echoes (and their SD) from Optison® diluted in saline at the first, second and subharmonic frequencies under different static pressures as a function of time. Transmitted frequency 4 MHz, 200 kPa (MI ⫽ 0.1), dissolved oxygen ⫽ 8.3 mg/L, water temperature ⫽ 23 °C. (a) 5 kPa, (b) 10 kPa, (c) 15 kPa and (d) 20 kPa.

682

Ultrasound in Medicine and Biology

Volume 31, Number 5, 2005

also tested: although Optison® deteriorates quite rapidly during static pressure (⬎ 10 kPa), the Optison® demonstrates somewhat better stability when slowly varying pressures are imposed and even better stability when “fast” (physiological) 1.0-Hz cyclic pressures are imposed. When such 1.0-Hz cyclic pressures are imposed, those with a range of pressures that start from 0 kPa to some higher pressure cause less deterioration than pressure changes from some high level (e.g., 10 kPa) to a higher level.

Fig. 9. The average and SD of the attenuation (in dB/cm) of Optison® diluted in saline under different static pressure (at 20 kPa, 15 kPA, 10 kPA and 5 kPa), as a function of time. Transmitted frequency 4 MHz, 200 kPa (MI ⫽ 0.1).

cluded that the overall changes of the attenuation have only a very minor effect on the results of the echo signals presented here. Brayman et al. (1996) used a similar set-up for testing the hypothesis that the acoustic transmittance of a UCA suspension increases with increasing pressure. Here, the assumption examined was that the acoustic attenuation, as well as the amplitude at the x0.5 harmonic, decrease with increasing ambient pressure. These two assumptions are basically equivalent. Brayman et al. (1996) confirmed their hypothesis for stable pressures and postulated that the first cycle of increase of pressure causes functional and irreversible affects on the UCA they studied, which was Albunex®. The results obtained here differ from the results mentioned above (Brayman et al. 1996) in several ways: ●





UCA behavior under “fast” (physiological) cyclic changes of pressure Very soon after the beginning of the experiment, when Optison® was diluted in the saline, a balance was reached at normal pressure by the air dissolved in the saline and the air that diffused into the Optison®. The inert gas (C3F8) dissolves very little in water; thus, it takes many 10s of minutes for this gas to diffuse out from the microbubbles. Because gas solubility is known to increase linearly with the magnitude of the applied pres-

When static ambient pressure was applied and when a specific time window was studied (the first min of US insonation), higher values of attenuation were obtained for Optison® UCA suspensions, after excessive pressure was applied. When varying pressures were applied, the results obtained here demonstrate that the pressure effect on the Optison® UCA behavior is reversible. When varying pressures were applied, good correlation was demonstrated between the behavior of the attenuation parameter and the changes in pressure.

The principal differences in these results can be explained by the different UCAs employed: Brayman et al. (1996) utilized Albunex® (albumin-shelled air bubbles) and, here, Optison® was employed, a UCA that contains inert gas, octafluoropropane (C3F8), instead of air and has an albumin shell with similar properties (Podell et al. 1999b). But also the pressure conditions tested were different. Here, cyclic pressure changes were

Fig. 10. Average attenuation (in dB/cm) of Optison® diluted in saline saturated with different gases vs. time (five experiments). Transmitting frequency of 1.8 MHz, MI ⫽ 0.1, PRF ⫽ 10 Hz. The attenuation curve is depicted for two static pressures, atmospheric and elevated pressure of 20 kPa. (a) Saline saturated with nitrogen; (b) with oxygen; and (c) with air; (d) degassed saline and (e) saline saturated with air, but the transmitted frequency here was 4 MHz, MI ⫽ 0.05.

Contrast agent and pressure ● D. ADAM et al.

Fig. 11. The changes in amplitudes of echoes from Optison® diluted in saline, at the subharmonic frequency (mean ⫾ SD) caused by cyclic pressure changes from the normal pressure, as a function of the amplitude of the pressure changes.

sure, as the pressure is increased, the capacity of the saline to dissolve air is increased and air moves out from the microbubbles back into the liquid. This process is reversed after the pressure is reduced back to normal pressure. Thus, when fast cyclic changes of pressure are imposed, they probably cause movement of air back and forth between the microbubbles and the liquid. During the 6-min experiment, while 1.0-Hz cyclic pressures are imposed, there is only a small change in the overall levels of the echo signals at the different harmonics (Fig. 2). The x1 harmonic and the x2 harmonic were very little affected in this case by the imposed pressure changes. The x0.5 harmonic, though, was very much affected, but it took some seconds for the response to develop (the full-scale fluctuations were reached after about 50 s). This result is clearly demonstrated when the correlation is calculated between the amplitudes of the harmonics and the imposed pressure; only the amplitude of the x0.5 harmonic is highly correlated to the pressure changes, but the correlation reaches a high value after 50 s, and the maximum value (⫺0.8) only after 100 s. The ratio of the second harmonic to half harmonic (as well as the ratio of the first harmonic to half harmonic) produces similarly clear signals (as well as the same levels of correlation to the imposed cyclic pressures), but the ratios provide values that are less dependent on the experimental conditions (e.g., concentration of UCA and US transmit levels). These experiments were made at different amplitudes of pressure changes (0 to 20 kPa, 0 to 15 kPa, 0 to 10 kPa and 0 to 5 kPa) that provide evidence that in vivo blood pressure (at least in the left ventricle) may be estimated from the measurements of the amplitudes of the echoes at the x0.5 harmonic of the transmitted frequency (Fig. 11). The correlation between these values and the small SD provide this evidence. When the attenuation is measured during the same experiments, it can be seen (Fig. 4) that the overall attenuation decays very little, about 1 dB/cm, when pressures up to 0 to 15 kPa are imposed. When higher

683

pressures are used, 0 to 20 kPa, the decay of the attenuation is of about 1.5 dB/cm during 6 min. The 1.0-Hz fluctuations of the attenuation caused by the imposed pressure changes are clearly seen in all of the experiments. The amplitudes of these fluctuations, though, hardly depend on the amplitudes of the imposed cyclic pressures; they are about 0.5 dB/cm after the pressure changes are above 0 to 10 kPa. The results are quite different when the cyclic pressures are imposed between a nonzero level and a higher level (e.g., 10 to 15 kPa or 15 to 20 kPa, see Fig. 6). Both the averaged amplitudes of the echoes at x0.5 and the attenuation demonstrate a significant decay already within the 6-min experiment. When cyclic pressures of 10 to 15 kPa are imposed, this decay of about 7 dB is clearly seen in the averaged amplitude of the echoes (Fig. 6a, left panel) after about 200 s, and the attenuation that slowly decays from the beginning of the experiment actually peaks at about 200 s, then decays at a higher rate (Fig. 6c, left panel). The 1.0-Hz fluctuations in the averaged amplitudes of the echoes and also the attenuation correlate well with the pressure changes, as seen in the right-hand panels of Fig. 6. When cyclic pressures of 15 to 20 kPa are imposed, the effects are more pronounced; the decay of the averaged amplitude of the echoes occurs already at 170 s (Fig. 6b, left panel) and the peak in the attenuation occurs at 140 s, with a total decay of 12 dB within 6 min. These results demonstrate that, in addition to the immediate response to the 1.0-Hz fluctuations, which are probably due to diffusion of air in and out of the microbubbles, other processes occur. These processes are hardly seen in the experiments of the 0 to 5 kPa, 0 to 10 kPa, 0 to 15 kPa and even the 0 to 20 kPa but, here, they are quite evident; the elevated pressure during the “diastolic” periods causes a more pronounced destruction of the UCA, as evidenced by the decay of the attenuation and, to some extent also, in the decay of the echoes. But, there is another process that is clearly seen only in the attenuation plots, which is an increase of the attenuation after some min of decay that causes a peak and then a more pronounced decay. This behavior may be explained by rectified diffusion, which is more pronounced when the “diastolic” pressures are elevated and which causes increase in size of the microbubbles, but also enhances their rate of destruction later. UCA behavior under “slow” cyclic changes of pressure Typical results of an experiment in which the cyclic pressure was imposed at a 0.05-Hz rate are given in Fig. 7. Excellent correlation is obtained between the cyclic pressure changes and the fluctuations of the attenuation parameter but, again, it takes some time for that to occur (about 50 s) After about 200 s, the attenuation decays nearly to zero and the amplitude of the fluctuations

684

Ultrasound in Medicine and Biology

becomes very small. One possible explanation is the destruction of the UCA. It is interesting to note, though, that the rate of decay is much faster than the rate when similar pressures were applied, but at the higher rate of 1.0 Hz. The rate of decay here is even faster than the 1.0-Hz cyclic pressure with elevated “diastolic” pressures (i.e., 10 to 15 kPa or 15 to 20 kPa, Fig. 6c and d). The rate here is slower than that during elevated static pressures (see below). The peak seen in the attenuation plots (Fig. 6c), left panel) and also during the elevated static pressures is not seen in these experiments. On the other hand, the increase in amplitude of the 0.05-Hz fluctuations (between 60 to 120 s) does indicate that the same process also occurs here (but without the increase of the transient low values of the attenuation when the pressure is at its maximal value). Again, it may point to an increase of inflow of air gases from the liquid to the microbubbles, in a process similar to “rectified diffusion” enhanced by the increase of pressure. This process occurs concurrently with the destruction of the microbubbles, which later causes the marked decay of the attenuation. It is interesting to note that, during the first few cycles of the pressure increase, the immediate response to the pressure increase is a transient increase of the attenuation. This increase of attenuation decays even within the 20 s of the elevated level of pressure. We are unable as yet to find an explanation for these results. UCA behavior under different constant elevated pressures When a constant elevated pressure of ⱕ 5 kPa is applied to Optison® diluted in gas-saturated saline, the average amplitudes of the echoes measured from the UCA at the transmitted frequency (x1), at the x2 harmonic and even at x0.5 harmonic remain nearly constant (⬍ 3-dB change) during the 20-min duration of the experiment (Fig. 8a). When a higher pressure is applied, the amplitudes of the echoes at x1 and at x2 deteriorate quite monotonously (by 20 to 25 dB for the x1 harmonic and 20 to 32 dB for the x2 harmonic); the amplitude of the x0.5 harmonic starts by deteriorating, but then the amplitudes significantly increase to produce a peak and then continue to deteriorate (by about a total of 15 dB). The peak is reached after some 6 to 7 min when 10 kPa is applied, after 4 to 5 min when 15 kPa is applied and after 3 min when 20 kPa is applied (Fig. 8b– d). The amplitude of the x0.5 is usually smaller than that of the other harmonics and the values are smaller when higher pressures are employed (Fig. 8a– d). The SDs of the amplitudes remain the same during the experiment with the 5 kPa and increase during the experiments of the higher pressures. The averaged attenuation measured concurrently

Volume 31, Number 5, 2005

also differs markedly between the results obtained when a constant elevated pressure of ⱕ 5 kPa is applied to Optison® diluted in gas-saturated saline and the results obtained when higher pressures are imposed. When 5-kPa pressure is imposed, the attenuation deteriorates linearly with time, with a reduction of about 1.2 dB/cm during the 20 min measured (Fig. 9). A higher static pressure causes the attenuation to deteriorate rapidly (similar to the results reported by others) (Brayman et al. 1996; Padial et al. 1995; Ran and Katz 1991) but, before the rapid decay, an increase occurs; when 10 kPa are imposed, there is a small decay, then an increase of the attenuation that peaks at about 2 to 2.5 min, then deteriorates very fast to half its value within the next min and to near zero by 15 min. When 15 kPa are imposed, the peak occurs earlier (⬇ 1.5 min), the decay is just as fast and it reaches near zero by 10 min. When 20 kPa are imposed, the peak occurs even earlier (⬇ 1.0 min); the decay, though, is less rapid, but still, within 10 min, it reaches near zero attenuation (Fig. 9). The decay (specifically when 10 kPa and 15 kPa are employed) is not exponential in nature; it does contain a second increase (that does not produce a peak) in attenuation. The SD of the averaged attenuation behaves very similarly to the average (Fig. 9). The SD is very high for the 10-kPa experiments, probably due to the low number of experiments (four) and slight inaccuracies in the experimental measurements. The diffusion processes, which have not been studied here, may explain the effect of lower attenuation values obtained when the higher excessive pressure was applied. The air gases that exist in the surrounding liquid will have a small effect on the microbubbles that contain air (Albunex®). But when microbubbles that contain a different, inert gas (e.g., C3F8 in the case of Optison®), are injected into the liquid, initially the process will be of the air gases (dissolved in the liquid) to enter the microbubbles. An attempt to explain the processes that occur later is made by performing the experiments described below. The difference in the behavior of the attenuation parameter between Optison® diluted in air-saturated saline and Optison® diluted in degassed saline, which was reported before, was also demonstrated here (Fig. 10c and d). The diffusion processes between microbubbles and surrounding liquid may explain this difference. In the degassed saline, small amounts of gases were dissolved and, thus, they did not diffuse into the microbubbles and, after enough octafluoropropane diffuses out from the microbubble, it may collapse. This represents itself as fast deterioration of the attenuation parameter (Fig. 10d). Thus, experiments were performed in which different gases were perfused through the degassed saline for more than 10 h so as to obtain

Contrast agent and pressure ● D. ADAM et al.

saline saturated with either nitrogen, Fig. 10a (because in in vivo blood only nitrogen is nearly at saturation level), oxygen, Fig. 10b, or air, Fig. 10c, transmitted frequency of 1.8 MHz, MI ⫽ 0.1, PRF ⫽ 10 Hz. Under normal room pressure, the decay of the US attenuation caused by the Optison® is quite similar, 0.4 to 0.9 dB/cm during 8 min of measurement, whether the saline is saturated with air, oxygen or nitrogen. When the pressure is increased by 20 kPa and maintained at this level, the attenuation first increases and peaks before it rapidly decays. The peak occurs somewhat earlier for air-saturated saline (⬇ 20 s), a bit later for oxygen (30 s) and a bit later for nitrogen. It is interesting to note that, although after the peak the decay is fast, after around 2 min, the decay changes its rate to a slower one. By 8 min, though, the Optison® under the 20 kPa pressure causes only very small attenuation (0 to 0.2 dB/cm). When the experiment was repeated with transmitted frequency of 4 Mhz, only for the Optison® in air-saturated saline, the results are quite similar (Figs. 9 and 10e), with less destruction of the microbubbles (Chomas et al. 2002). The different pressure conditions used in this study provide some insight into the different processes that occur when the Optison® microbubbles are injected into liquid containing air gases. The very initial response is due to the gases entering the microbubbles. When the ambient pressure increases, the amount of gases diffused from the surrounding liquid into the bubble also increases. The measurements here were done in vitro with saline as the surrounding liquid, not blood. The pressure changes were done at a rate of 1.0 Hz, with results being more accurate and repeatable than those obtained at slower rates (e.g., at 0.05 Hz), or under static pressures. Over the time period measured, Optison® demonstrated minimal deterioration when 1.0-Hz cyclic pressure changes were used. The effects found here will probably be similar within the body of a subject to whom a second-generation UCA (e.g., Optison®) is being injected. The blood contains dissolved gases and the pressure changes will affect the diffusion processes. But the conditions seem favorable, so that the pressure changes may be measured through the changes in the UCA properties and produce a noninvasive estimate of relative pressures. Acknowledgements—The authors thank the continuous support of the Technion VPR Fund for the promotion of research. The authors are also grateful to the IZMEL consortium, which supported, in part, this study.

REFERENCES Bouakaz A, de Jong N, Cachard C, Jouini, K. On the effect of lung filtering and cardiac pressure on standard properties of ultrasound contrast agent. Ultrasonics 1998;36:703–708.

685

Bouakaz A, Frinking PJ, de Jong N, Bom N. Noninvasive measurement of the hydrostatic pressure in a fluid-filled cavity based on the disappearance time of micrometer-sized free gas bubbles. Ultrasound Med Biol 1999;25:1407–1415. Brayman AA, Azadniv M, Miller MW, Meltzer MW. Effect of static pressure on acoustic transmittance of Albunex microbubble suspensions. J Acoust Soc Am 1996;99:2403–2408. Chen WH, Shung KK. The effect of low intensity ultrasound on Optison. Proc IEEE Ultrason Sympos 1998;2:1791–1794 Chen Q, Zagzebski J, Wilson T, Stiles T. Pressure-dependent attenuation in ultrasound contrast agents. Ultrasound Med Biol 2002;28: 1041–1051. Chomas J, Dayton P, May D, Ferrara K. Nondestructive subharmonic imaging. IEEE Trans Ultrason Ferroelec Freq Control 2002;49: 883– 892. de Jong N, Ten Gate FJ, Vletter WB, Roelandt JR. Quantification of transpulmonary echocontrast effects. Ultrasound Med Biol 1993; 19:279 –288. de Jong N, Frinking PJ, Bouakaz A, et al. Optical imaging of contrast agent microbubbles in an ultrasound field with a 100-MHz camera. Ultrasound Med Biol 2000;26:487– 492. Fairbank WM, Scully MO. A new noninvasive technique for cardiac pressure measurement: Resonant scattering of ultrasound from bubbles. IEEE Trans Biomed Eng 1977;24:107–110. Frinking PJ, Cespedes EL, Kirkhorn J, Torp HG, de Jong N. A new ultrasound contrast imaging approach based on the combination of multiple imaging pulses and a separate release burst. IEEE Trans Ultrason Ferroelec Freq Control 2001;48: 643– 651. Greim CA, Broscheit JA, Lorenz KW, Thiel H, Roewer N. Intracavity contrast intensity after transpulmonary transmission of a second-generation contrast agent at normal and reduced myocardial contractility. J Am Soc Echocardiol 2000;13:1030 – 1037. Hok B. A new approach to noninvasive manometry: interaction between ultrasound and bubbles. Med Biol Eng Comput 1981;19:35–39. Hynynen K, McDannold N, Martin H, Jolesz FA, Vykhodtseva N. The threshold for brain damage in rabbits induced by bursts of ultrasound in the presence of an ultrasound contrast agent (Optison®). Ultrasound Med Biol 2003;29:473– 481. Kabalnov A, Bradley JA, Flaim S, et al. Dissolution of multicomponent microbubbles in the bloodstream: 2. Experiment. Ultrasound Med Biol 1998a;24:751–760. Kabalnov A, Klein D, Pelura T, Schutt E, Weers J. Dissolution of multicomponent microbubbles in the bloodstream: 1. Theory. Ultrasound Med Biol 1998b;24:739 –749. Kirkhorn J, Frinking PJ, de Jong N, Torp HG. Three-stage approach to ultrasound contrast detection. IEEE Trans Ultrason Ferroelec Freq Control 2001;48:1013–1022. Miwa H. Pressure measurement system with ultrasonic wave. US patent no. 4,483,345. 1984. Padial LR, Chen MH, Vuille C, et al. Pulsatile pressure affects the disappearance of echocardiographic contrast agents. J Am Soc Echocardiol 1995;8:285–292. Podell S, Burrascano C, Gaal M, et al. Physical and biochemical stability of Optison®, an injectable ultrasound contrast agent. Biotechnol Appl Biochem 1999a;30:213–223. Podell S, Golec B, Lohrmann R. Measuring the effects of ultrasound on contrast agents, 1999 IEEE Ultrasonics Symposium. Piscataway, NJ: IEEE, 1999b:1749 –1754. Ran B, Katz J. The response of microscopic bubbles to sudden changes in the ambient pressure. J Fluid Mech 1991;224:91–115. Reddy AK, Taffet GE, Madala S, et al. Noninvasive blood pressure measurement in mice using pulsed Doppler ultrasound. Ultrasound Med Biol 2003;29:379 –385. Shankar PM, Chapelon JY, Newhouse VL. Fluid pressure measurement using bubbles insonified by two frequencies. Ultrasonics 1986;24: 333–336.

686

Ultrasound in Medicine and Biology

Shankar PM, Krishna PD, Newhouse VL. Subharmonic backscattering from ultrasound contrast agents. J. Acoust Soc Am 1999;106: 2104 –2110. Shapiro JR, Reisner SA, Lichtenberg GS, Meltzer RS. Intravenous contrast echocardiography with use of sonicated Albumin in humans: Systolic disappearance of left ventricular contrast after transpulmonary transmission. J Am Coll Cardiol 1990;16:1603–1607. Shi WT, Forsberg F. Ultrasonic characterization of the nonlinear properties of contrast microbubbles. Ultrasound Med Biol 2000;26:93–104. Shi WT, Forsberg F, Hall AL, et al. Subharmonic imaging with microbubble contrast agents: Initial results. Ultrason Imag 1999a; 21:79 –94.

Volume 31, Number 5, 2005 Shi WT, Forsberg F, Raichlen JS, Needleman L, Goldberg BB. Pressure dependence of the subharmonic signals from contrast microbubbles. Ultrasound Med Biol 1999b;25:275–283. Strauss AL, Roth FJ, Rioger H. Noninvasive assessment of pressure gradients across iliac artery stenoses: Duplex and catheter correlative study. J Ultrasound Med 1993;12:17–22. Van Liew HD, Burkard ME. Behavior of bubbles of slowly permeating gas used for ultrasonic imaging contrast. Invest Radiol 1995;30: 315–321. Van Liew HD, Raychaudhuri S. Stabilized bubbles in the body: Pressure-radius relationships and the limits to stabilization. J Appl Physiol 1997;82:2045–2053.