Noninvasive temperature estimation technique for HIFU therapy monitoring using backscattered ultrasound

Abstract

High Intensity Focused Ultrasound (HIFU) has potential applications in a variety of medical fields that benefit from selective destruction of tissue volumes. Being a noninvasive technique, HIFU provides the ability to ablate localized tissue volumes without damaging intervening tissue, thus eliminating the need for incisions. However, the lack of real-time feedback during the various stages of therapy has been a major hindrance to its widespread clinical acceptance. The ability to accurately track and monitor the therapy progress is therefore important for the success of HIFU treatment protocols.This work consists of three specific goals. First, an integrated software-based experimental data acquisition system was built, using a commercially available ultrasound scanner, to collect the raw ultrasound backscatter (RF) during HIFU therapy. Second, signal processing approaches were developed to visualize the temporal evolution of lesion formation by analysis of the RF data. Temperature change related information throughout HIFU therapy delivery and post-treatment cool-down was obtained by tracking echo shifts in the RF using cross-correlation techniques. Characteristic changes in the spectral content of RF echoes, due to changes in the scattering properties of the heated region, were observed well before the appearance of hyper-echogenic regions on B-mode images in the focal zone. Third, a bioheat transfer equation model-based non-invasive quantitative temperature estimation technique for HIFU therapy monitoring was developed and validated in tissue mimicking phantoms and excised animal tissue. In this approach, initial estimates of local thermal and acoustic parameters are first obtained prior to therapy through "calibration" exposures performed in the treatment region. Tissue heterogeneity over a larger targeted region of interest is modeled as a change in the magnitude of the heating rate. During therapy, this magnitude is updated using an optimization technique that minimizes the difference between the model predicted and measured echo shift values. Ultrasound derived temperature estimates were validated against independent thermocouple measurements, close to but not at the HIFU focus. This model-based technique permits noninvasive temperature estimation over the entire therapeutic range, and is thus a departure from previously reported ultrasound-based temperature estimation techniques. These results demonstrate potential for applicability of the techniques in image-guided HIFU therapy.