Tissue motion assessment from 3D echographic speckle tracking

Abstract

The potential of ultrasonic image speckle tracking to characterize tissue dynamics has been illustrated and validated elsewhere. In this paper we wish to extend this speckle tracking methodology to 3D. To investigate the feasibility of such an approach we first model the image formation process and simulate the 3D speckle motion inherent to tissue linear transformations (translation, rotation and deformation). It is shown that tissue axial rotation and translation are perfectly correlated with the tissue speckle motion while tissue deformation and non-axial rotations corrupt the speckle pattern with a motion-induced noise and are therefore more difficult to track when large motions are concerned. Furthermore, in the framework of our model, our results indicate that short ultrasound pulses with low frequencies and small beamwidths are more desirable for a speckle tracking methodology. The feasibility of speckle tracking is illustrated with an optical flow algorithm. A theoretical study of the correlation between various linear transformations of the tissue and the corresponding ultrasonic speckle motions is also performed.