Various methods have been proposed to image elastic properties of soft tissues. This imaging takes advantage of changes on elastic properties that are markedly different for some abnormal structures such as malignant tissues, ablation lesions or calcifications from surrounding tissues. In particular, ultrasound-based elasticity imaging has the potential of being a powerful low-cost monitoring tool for high intensity focused ultrasound (HIFU) thanks to large changes in elastic properties for coagulated tissues.
Elastic properties of thermally coagulated tissue induced by HIFU may be visualized by different approaches. Acoustic radiation force impulse imaging (ARFI) uses a localized, impulsive radiation force to excite the target tissues and the generated displacements are used to evaluate stiffness. Ultrasound-stimulated acoustic emission (USAE) employs the superposition of ultrasound beams to exert a varying force on the tissues and the emission field is used to form an image. Elastography produces strain images from displacement calculations while continuously compressing tissues. Supersonic shear imaging (SSI) uses ultrasonic focused beams to generate shear waves inside tissues and then mapping the shear elasticity. Localized harmonic motion imaging (LHMI) uses an amplitude-modulated focused ultrasound exposure to induce a time-varying force oscillation to generate a motion at the focus. In LHMI this motion is done concurrently with focused ultrasound energy, and the amplitude at the focus is affected by the changes induced to tissues during this ablation, making it possible to track lesion formation and monitor treatment.
The mechanism behind all these techniques is the induction of displacement by different means and following up through the radio-frequency (RF) signals coming from ultrasound echoes acquired during this displacement. Tissue motion is typically estimated at different time points using cross-correlation of the signal at a specific time point vs. a reference signal, where this reference can be the initial signal before displacement or the previous time point. The echoes are usually broken into small overlapping segments and cross-correlation is applied between congruent segments in a line pair with the maximum peak of the cross-correlation used as the estimate of the time shift. If the elastic modulus differs somewhere along the line, little increase will show in the time shift of certain segments. Finally, an image of tissue properties is obtained by using multiple sets of line pairs.
Multiple challenges are faced by most ultrasound imaging of elastic properties: 1) the physical displacement of tissues, depending on the technique used to induce displacement, may be complex because of the inhomogeneity of tissues; 2) it is often hypothesized that the displacement is limited to the same line, but structures may move out of the line and the motion in that direction is not tracked; 3) in order to achieve accurate motion tracking high sampling rate of the RF signal is used, increasing the computational requirements and making real-time imaging challenging; 4) most ultrasound scanner systems are now using arrays where beam-forming techniques most be applied before the cross-correlation is applied, adding to the computational challenges for obtaining the final maps.
This is of particular concern when using LHMI as a targeting and monitoring technique since tissues are being altered by coagulation and fast detection of changes is required for treatment safety and accuracy of the imaging. In addition, the high intensity ultrasound waves cause a challenging noisy environment for acquisition of ultrasound echoes, requiring the implementation of fast and efficient digital notch filtering that should not alter phase or signal signature to the point that the cross-correlation is affected.
In this presentation a review of LHM imaging and how it was validated in phantoms and animals for targeting, monitoring and imaging is presented. The particular challenges of the acquisition and processing of the signals and the current developments using ultrasound imaging arrays and 2D imaging is presented. The objective is to discuss opportunities on possible tools that can be applied to this problem to achieve real-time and computationally low-cost processing of this signals to achieve reliable images for monitoring and targeting focused ultrasound lesions with ultrasound techniques.