Dr. Schwarz is an Associate Professor of Medicine and the Director of the Echocardiography Laboratory at University of Rochester Medical Center, Rochester, NY.

Microbubble echo contrast is well-documented to improve left ventricular endocardial border delineation in patients with technically limited examinations, especially when harmonic imaging is used. However, it is not uncommon for some patients with very technically limited exams to have less than complete endocardial visualization, even with the use of harmonics and echo contrast. Such may often be the case for stress echo patients in the immediate post-stress period. Harmonic tissue Doppler energy (TDE) combines the motion filter of tissue Doppler (which rejects non-moving image artifacts) with the image improving effects of harmonic imaging (higher resolution imaging and near elimination of side lobe artifact compared to fundamental frequency imaging) and will often improve image resolution and dramatically reduce image clutter. In addition, high-power tissue Doppler ultrasound enhances microbubble imaging through bubble destruction, which creates broadband ultrasound transients and results in loss of correlation (for the Doppler autocorrelator) and a stronger than normal Doppler signal. Combining harmonic TDE as a semi-transparent overlay to the two-dimensional (2D) native tissue harmonics image (NTHI) maximizes the contrast enhancement of the blood-pool relative to the myocardial tissue. In this way, the effects of Harmonic TDE and the NTHI are additive to improve endocardial border delineation, while at the same time not increasing image clutter. Image frame rates of at least 30 Hz are possible using ultrasound scanners containing a coherent beamformer, thus allowing real-time Harmonic TDE imaging of endocardial motion with similar temporal resolution to traditional 2D NTHI.

Background

The sources of image clutter, and how NTHI and TDE imaging improve ultrasound images in technically limited exams with a low signal-to-noise ratio, will be discussed in detail in a forthcoming clinical white paper.

Echo Contrast Mini Bubble Physics Review

All echo contrast agents currently available consist of very small microbubbles delivered intravenously in an aqueous medium. These microbubbles initially exist entirely in the blood-pool and enhance ultrasound backscatter through two primary mechanisms: 1) speed of sound change at the gas-fluid interface of the microbubble; and 2) though micro-bubble oscillation, which increases the effective cross-section of the micro-bubble and also radiates ultrasound energy. The optimum steady-state backscatter is achieved from a microbubble when it is insonified at its resonance frequency with optimal power settings. The resonance frequency is based on a number of factors, but the most important are the bubble size and the insonification frequency. The resonance frequency can be roughly calculated as the dividend of 6 and the bubble diameter. At standard clinical harmonic frequencies (transmit 2 to 1.75 MHz), this corresponds to bubble diameters of 3 to 3.5 µm. The echo contrast agents currently available for clinical use consist of a population of bubble sizes, but the population means of bubble diameters are in this range. A higher transient spike of ultrasound backscatter may also be elicited from bubbles when they are exposed to higher power ultrasound and there is microbubble destruction. Thus, optimal steady-state contrast imaging may be achieved at the resonance frequency using modest insonification energies, but maximal contrast imaging will occur using high-power insonification energies at the resonance frequency. High-power insonification energies destroy the bubbles as they are being imaged, so rapid replenishment of contrast is required to achieve continuous imaging.

How Echo Contrast Improves Left Ventricular (LV) Endocardial Border Delineation

Imaging of the LV endocardial borders are limited by factors that attenuate the ultrasound beam (obesity, narrow intercostals spaces, hyperinflated lungs in pulmonary disease patients, etc.) and cause cardiac displacement (as in obese patients and those with pulmonary disease). LV endocardial visualization is also limited by the lateral resolution of the ultrasound system, because much of the endocardial border is parallel to the ultrasound beam when imaging from the apical window. Microbubbles are the solution to these problems based in part on their shape (spherical microbubbles are always imaged with axial resolution) and by their enhanced backscatter (especially at the harmonic frequencies) compared with the non-bubble surrounding medium. When given in sufficient quantities, the borders of the microbubble-rich blood-pool define the endocardial border.

How 2D Harmonic Imaging Improves Microbubble Imaging

Second harmonic imaging improves the imaging of microbubble echo contrast in part the same way it improves NTHI. Compared with fundamental frequency 2D imaging, harmonic imaging uses relatively high receive frequencies and a more narrow transmit beam (better image resolution) and NTHI transmit beams have less side-lobe content (less image clutter). Microbubble backscatter power is greatest at the fundamental frequency. However, the relative backscatter power of microbubbles compared with tissue is much greater at the second harmonic frequency than at the insonification frequency (Figure 1). This leads to improved border delineation and better detection of low concentration microbubbles, when using second harmonic imaging (Figure 2). In addition to these factors, high-intensity imaging produces violet microbubble oscillations and sudden bubble destruction, which produces transients of even greater harmonic backscatter.

How TDE Reduces Improves Microbubble Imaging

TDE imaging improves NTHI imaging for the detection of endocardial borders through the addition of the Doppler motion filter (removes unwanted non-moving image "clutter"). As noted above, microbubbles that are oscillating violently and are destroyed in high intensity sound fields transmit high-intensity broadband transient ultrasound bursts. The Doppler auto-correlation algorithm interprets the ultrasound transients associated with bubble destruction as motion, usually of uncertain velocity, but of high intensity. This effect typically creates a vivid color Doppler image with marked "aliasing" noted, even if there is no blood flow velocity. The tissue Doppler energy map also produces a vivid image, but it appears uniform because the energy map is devoid of velocity information. The endocardial border is identified as the border zone between the bubble-rich blood-pool and the bubble-poor myocardium. TDE has much higher frame rates than traditional color Doppler due to a smaller packet size, especially when a coherent beamformer is used. This allows TDE imaging of all but the largest ventricles at a frame of 30 Hz or more.

Combined NTHI and TDE for Microbubble Imaging

TDE images are displayed as a semitransparent overlay to the NTHI image. Both TDE and NTHI enhance the bubble-rich blood-pool relative to the bubble-poor myocardium. Therefore, bubble-rich backscatter is enhanced when NTHI and TDE are displayed as an additive image, when compared with displaying either singularly (Figure 3).

Instrument Settings for NTHI/TDE

The goals of the instrument settings for NTHI and TDE are to maximize the contrast-rich intracavitary blood-pool backscatter, while at the same time minimizing the bubble-poor myocardial backscatter. Stated another way, the left ventricular cavity should appear "white" while the myocardium should appear "black." This is exactly the opposite of the normal goals when imaging the left ventricle without contrast; normally the machine is adjusted so that the myocardium is white and the cavity is black. As noted above, both contrast and tissue backscatter intensity is similar when imaged at the insonification frequency (Figure 3). This leads to a "white" left ventricular cavity and a "white" myocardium (poor differentiation between contrast and tissue). However, tissue backscatter is much lower in intensity relative to contrast bubbles at the second harmonic, and this difference can be accentuated if low transmit power settings are used. This is due to the nonlinear relationship between transmit power and NTHI backscatter; for every drop in transmit power, there is a correspondingly larger drop in NTHI backscatter. In contrast, microbubble resonance still occurs at lower power, which accounts for the much greater backscatter intensity of microbubbles relative to tissue at low transmit powers when using harmonic receivers.

The instrument settings for NTHI and TDE are best adjusted separately, usually before the introduction of echo contrast. Each patient will require minor adjustments to the settings as described above, depending on the imaging circumstances. Occasionally, the settings will need to be adjusted during the contrast infusion. In the case of stress echocardiography, the pre-stress image settings usually determine the post-stress settings and it is seldom necessary to make adjustments to the instrument in the post-stress period.

Contrast 2D Harmonic Settings (Table 1, Figures 4A and 4C)

Normally maximal transmit power is used for noncontrast NTHI to maximize the amount of nonlinear wave propagation and thus maximize NTHI backscatter. The 2D harmonic settings are slightly different when contrast is employed compared with harmonic imaging used without contrast. The main difference is that the output power is adjusted downward to maximize bubble backscatter relative to tissue and to minimize the bubble destruction associated with high output power. Typically, transmit power settings of ¾0.8 MI are used for harmonic contrast imaging. As is the case in NTHI imaging, the transmit focus should be placed at the area of interest for small targets or near the far-field boundary for larger targets (ie, near the mitral valve plane, when imaging the LV from the apical window). The image dynamic range needs to be set at a level that places the contrast bubble echoes in the middle of the image gray-scale, while at the same time minimizes the display of myocardial echoes. This is usually achieved with a dynamic range setting of about 60 dB, with the image (log) compression curve chosen to further suppress low intensity (myocardial) backscatter and image noise. The optimum gain setting is one in which endocardial borders are just barely visible on the pre-contrast images and without excess clutter artifact. This may mean that the NTHI image looks quite light (under-gained). Regardless of the actual overall gain setting, it is critical
that the depth-dependent time-gain (TGC) slider controls be used for adjusting gain first, with overall system gain being used only for final adjustments. In all cases, the TGC slider controls should be somewhere in the 50% to 75% gain positions with the overall system gain used to bring the NTHI image either up or down in gain to the optimal range.

Contrast Harmonic TDE Settings (Table 1, Figures 4B and 4D)

The instrument is set to color TDE imaging and adjusted so that the imaging frequency is the same as that used for NTHI imaging (typically transmit 1.75 MHz, receive 3.5 MHz). The frequency adjustment is important so that the instrument can be optimized in both the NTHI and TDE modes for simultaneous use. The transmit TDE power of the machine is set to maximum, so that there will be maximal microbubble oscillation, including some microbubble destruction, which will enhance the color Doppler signal. The color filter should be set to give good sensitivity while removing stationary "clutter" from the image (filter 2 on the Acuson products). The harmonic color TDE dynamic range is typically adjusted to a level about the same as the NTHI image, but usually not more than 60 dB and not less than 50 dB in very technically limited cases. The NTHI image then needs to be turned off temporarily, leaving just the harmonic color TDE image. Assuming the machine has already been setup for optimal NTHI imaging (TGC slider controls in the 50% to 75% gain positions), Doppler gain is then adjusted to just bring out endocardial borders without excessive image artifact. This gain setting is less than typical for noncontrast imaging so that the system will not be over-gained once contrast is added to the image. The image frame-rate is optimized through a combination of reducing 2D harmonic/TDE image width, and by adjusting TDE packet-size and line-density ("Space-Time" control to T2 on the Acuson products). A frame-rate of at least 30 Hz should be achievable in nearly all cases when using an instrument with a coherent beamformer. The TDE image is then changed back to the semi-transparent overlay with the NTHI image (mix = 3 or 4 on the Acuson products) and a pleasing color map is chosen. While any color map will suffice, a gray-scale map is often used so that the combination 2D harmonic/harmonic color TDE image appears no different from a more traditional 2D NTHI image.

Combined 2D Harmonic/TDE Settings

Echo contrast should appear as a very bright substance that fills the entire LV cavity while the myocardium looks quite dark, thus allowing for the identification of the endocardial boundaries. No adjustments of the instrument settings should be required during contrast imaging, if the machine has been setup as described above. It is not uncommon for "swirling" of the contrast to be observed in the near field of the ultrasound image, where the ultrasound intensities are highest. The "swirling" is due to a combination of microbubble destruction and movement of the bubbles due to acoustic radiation force, and usually most obvious in the LV apex when imaging from the apical window. The appearance of "swirling" is dependent on the ultrasound transmit intensity, the fragility of the microbubbles, and on the rate of microbubble replenishment in the imaging area. It is for this reason that low transmit intensities are used for 2D harmonic contrast imaging, but high transmit power is required for TDE imaging. Under resting conditions, the rate of contrast replenishment in the imaging area is dependent on the rate or dose of contrast administered. Therefore, it is typical to use a higher rate of contrast infusion when using TDE imaging compared with 2D harmonic imaging alone. In the case of peak or post-peak stress imaging, the rate of contrast replenishment is using high enough to avoid "swirling" due to the high cardiac output and no adjustment to the contrast infusion rate need be made.

Potential Uses for Contrast 2D Harmonic/TDE Imaging

Contrast 2D harmonic/TDE imaging can be used in all settings that might benefit from contrast enhancement of the blood-pool, regardless of the pre-contrast image quality. While the addition of TDE results in an improved border delineation between the bubble-rich blood-pool and tissue compared to 2D harmonic imaging in patients with high-quality acoustic windows, the benefits of TDE are most apparent in patients with low-quality acoustic windows. In this setting, the addition of the TDE motion filter combined with the high-intensity acoustic transients associated with high transmit power induced microbubble destruction, results in greatly improved signal-to-noise ratio compared to 2D harmonics alone. It is not uncommon to have diagnostic contrast 2D harmonic/TDE images in patients with nondiagnostic contrast 2D harmonic imaging used alone. Therefore, the target patient population most likely to benefit from the combined imaging approach is patients with very limited noncontrast 2D harmonic images that need resting or stress evaluation of LV systolic function. *