Quantitative analysis and applications of non-invasive temperature estimation using diagnostic ultrasound

A quantitative analysis of a two-dimensional non-invasive real-time temperature-change estimation algorithm based on diagnostic backscattered ultrasound RF-data is presented in this paper. An experimental system consisting of a therapeutic ultrasound unit and an ultrasound imaging unit was used to quantitatively characterize the accuracy, spatial resolution and ripple artifacts of the temperature estimates in a rubber phantom heating experiment. The ripple is shown to be a consequence of the thermo-acoustic lens effect resulting from local changes of the speed of sound in heated regions. Non-invasive temperature estimates were used as inputs to a multipoint ultrasound phased array temperature controller, for a long duration hyperthermia experiment. The applicability of this method to tissue media, as well as its major limitations are discussed.