Dose uniformity in scanned focused ultrasound hyperthermia

Abstract

Scanned focused ultrasound (SFUS) is unique amongst noninvasive methods of inducing hyperthermia in that the absorbed power (SAR) distribution may be controlled at a scale of 0·5 cm or better. This high degree of spatial control of SAR implies that differences in local tissue cooling due to heterogeneity in perfusion, variations in the density of discrete thermally significant vessels and even local cooling around single large vessels may be compensated during SFUS treatments. In this paper we calculate thermal dose distributions arising from three different SFUS techniques: (1) the high-temperature short-duration (HTSD) ultrasound technique; (2) conventional systems employing a fixed scan and fixed SAR simulated by a uniform SAR; and (3) we optimize the SAR distribution using information from angiography and thermal models. These techniques are tested in the same anatomy having discrete vessels, a non-uniform vessel density and with little, if any, preheating of the incoming blood (i.e. the worst-case situation). The application of a uniform SAR to this volume resulted in a highly non-uniform thermal dose distribution. The situation is clearly improved in HTSD hyperthermia; however, the desired accuracy of positioning within the tumour is high, and to minimize background heating the focus must be stepped slowly and preferably in a semi-random pattern through the target volume. The third technique which employs spatial control over SAR resulted in the greatest uniformity in thermal dose. However, the use of this technique requires the input of the complete three-dimensional discrete vessel network and the availability of a tested three-dimensional discrete vessel thermal model.