Temperature estimation using ultrasonic spatial compound imaging

The feasibility of temperature estimation during high-intensity focused ultrasound therapy using pulse-echo diagnostic ultrasound data has been demonstrated. This method is based upon the measurement of thermally-induced modifications in backscattered RF echoes due to thermal expansion and local changes in the speed Of Sound. It has been shown that strong ripple artifacts due to the thermo-acoustic lens effect severely corrupt the temperature estimates behind the heated region. We propose here a new imaging technique that improves the temperature estimation behind the heated region and reduces the variance of the temperature estimates in the entire image. We replaced the conventional beamforming on transmit with multiple steered plane wave insonifications using several subapertures. A two-dimensional temperature map is estimated from axial displacement maps between consecutive RF images of identically steered plane wave insonifications. Temperature estimation is then improved by averaging the two-dimensional maps from the multiple steered plane wave insonifications. Experiments were conducted in a tissue-mimicking gelatin-based phantom and in fresh bovine liver.

I. Introduction

The feasibility of temperature estimation during high intensity focused ultrasound (HIFU) therapy using pulse-echo diagnostic ultrasound data has been demonstrated by Seip et al. [1], [2]. HIFU is a promising therapeutic technique for the treatment of deep localized cancers [3], [4]. However, improved monitoring methods are necessary before HIFU can be clinically useful. Accurate treatments specifically require that temperature elevations in the treatment volume be controlled with appropriate feedback. Over the past decade, several research groups have focused on non-invasive temperature estimation using magnetic resonance imaging [5]–[7] and ultrasonic imaging [2], [8], [9]–[12]. Both of these techniques have their own advantages and limitations. Magnetic resonance imaging of temperature provides quasi-real time three-dimensional temperature maps with a very good localization of the heated region, but is costly to use. In contrast, ultrasound techniques can be easily integrated into HIFU systems with relatively low cost and a high portability.