Evaluation of Tissue Motion Vector Distribution Utilizing Synthetic Aperture Array-Signal Processing Triggered with Successive Virtual Source Generation

摘要

The real-time elastography [1]-[5] is expected as a promising breakthrough of conventional static elastography, which visualizes viscoelastic parameters and their dynamic behaviors in living soft tissue for diagnostic and therapeutic clinical approaches in vivo. For the dynamic imaging of quantitative shear elasticity the accurate and high-speed measurement of spatio-temporal displacement vector across tissue parenchyma have to be required to separate shear wave component obtained by accounting vector displacement potential under internally physiological or externally imposed stress field. Thus our proposed breakthrough came up with the combined algorithm of conventional cross-correlation technique and synthetic aperture (SA) array-signal processing of successive speckle echo frame. The orthogonal components of local displacement vector can be uniquely determined by spatially separated and temporally successive irradiations from transducer controlled virtual sources. The performance of this proposed method was 2-dimensionally analyzed by utilizing a tissue medium simulated as randomly distributed point scatterers with gaussian reflectivity in front of an array transducer 64 millimeters wide with half wavelength element pitch at a center frequency of 3 MHz. For the longitudinal and lateral displacement of 0.05 mm the excellent variances less than 2.7% and 12.1% in standard deviation to displacement ratio were obtained respectively, along the longitudinal distance from virtual sources to the tissue origin (52 mm to 86 mm) around the array position (69 mm). It was further shown that the rotational motion vectors reconstructed by the spatial averaging of local positional evaluation between two echo frames were rather smooth than by the ordinary cross-correlation for a slight tissue rotation by 0.3 degree. Experimental approaches also verified the simulated performance of the proposed method by utilizing a customized linear array transducer and 256 channel AD converters with static memories.