(E-163) Minimizing Uncertainty in Tissue Acoustic Property Measurements

High-intensity focused ultrasound (HIFU) uses sound waves to non-invasively ablate tumors deep within the body. An ultrasound transducer focuses the sound waves like a magnifying glass focuses light, and due to the high concentration of energy, quickly generates enough heat to destroy tissue in a very small region. Computer simulations of HIFU can be used for optimal treatment planning, for improved monitoring of treatments, and for guiding real-time predictive controllers. These simulations require inputs of speed of sound and attenuation. These acoustic properties help predict the location and intensity of the focus, as well as energy deposited in other locations. Accurate acoustic property measurements lead to accurate simulations, which are important for effectively ablating the tumor and ensuring the safety of healthy tissues. The purpose of this study is to identify experimental conditions that result in the most consistent acoustic property measurements for HIFU simulations.

This study included two acoustic property measurement setups with a gelatin phantom serving as a tissue surrogate.

First, the through-transmission setup utilized the insertion-loss technique with a planar ultrasound transducer (Olympus, Model #V314-SU, diameter = 0.75”) and hydrophone (ONDA, Model HNR-1000) to characterize the signal shift (linked to speed of sound) and decay (linked to acoustic attenuation) when the gelatin phantom was placed between them [Left image of Figure 1]. The through-transmission setup measured speed of sound and acoustic attenuation at multiple frequencies (Range = 0.6-3.0 MHz) with varied experimental conditions such as distance between the tip of the hydrophone and transducer (Range = 13-26 cm).

The second setup, called radiation force balance, used the insertion-loss technique to characterize acoustic attenuation.  A single element focused ultrasound transducer (Sonic Concepts, Model H-104, diameter = 64 mm, focal length = 63.2 mm) produced waves vertically in a water bath, and an acoustically absorbing target connected to a scale [Right image of Figure 1] measured the radiation force exerted by the ultrasound. The reduced scale reading when the gelatin phantom was placed between the transducer and target was used to compute the phantom’s acoustic attenuation. Radiation force balance measurements were conducted at the transducer’s fundamental and third harmonic frequencies (476 kHz and 1597 kHz). Variable experimental conditions included the acoustic power (10-50 W) and the position of the gelatin phantom relative to the transducer’s geometric focus (focus above, inside, and below the phantom).

In the through-transmission setup, careful and consistent windowing of the primary signal reduced errors. Further, accurately measuring phantom thickness was necessary to accurately quantify speed of sound. Increased water temperature was observed to increase speed of sound measurements, suggesting that water temperature ought to be a controlled experimental variable.

When investigating the effect of distance between the transducer and hydrophone, all speed of sound values were reasonable, but there was greater variation as the distance increased [Figure 2]. Attenuation measurements at 19.2 cm and 25.9 cm resulted in non-physical negative values, while measurements at 13.0 and 16.7 cm were consistent with those in the literature (Farrer et al. J Ther Ultrasound 2015). Further, there was greater variation in acoustic attenuation as the distance increased. Therefore, we recommend using less distance between the transducer and hydrophone for through-transmission measurements.

In the radiation force balance setup, noise in the scale readings had a greater impact at lower powers [Figure 3]. Therefore, we recommend using powers at or above 20 W.

Observationally, our initial radiation force balance data was the most irregular, and our process improved with time and experience. We recommend performing practice experiments before recording measurements where you want reliable and repeatable results.

Ultrasound physics and the radiation force balance technique provide two rules that guided our assessment of the measurements in Figure 4.  First, acoustic attenuation at the fundamental frequency should be lower than at the third harmonic frequency. Additionally, at a given frequency and focus position, attenuation should be independent of the acoustic power. Based on these rules and the data in Figure 4, we recommend placing the focus either above or within the gelatin phantom, rather than below.

In the big picture, there will always be some uncertainty in HIFU treatments due to patient-to-patient variability of acoustic properties and tissue heterogeneity. However, applying the recommendations above will improve the accuracy of predictive HIFU simulations by reducing human and experimental process error in acoustic property measurements.