Three-dimensional shear wave elastography on conventional ultrasound scanners with external vibration.

2020 
Two-dimensional (2-D) ultrasound shear wave elastography (SWE) has been widely used for soft tissue properties assessment. Given that shear waves propagate in three dimensions (3-D), extending SWE from 2-D to 3-D is important for comprehensive and accurate stiffness measurement. However, implementation of 3-D SWE on a conventional ultrasound scanner is challenging due to the low volume rate (tens of Hertz) associated with limited parallel receive capability of the scanner's hardware beamformer. Therefore, we developed an external mechanical vibration-based 3-D SWE technique allowing robust 3-D shear wave tracking and speed reconstruction for conventional scanners. The aliased shear wave signal detected with a sub-Nyquist sampling frequency was corrected by leveraging the cyclic nature of the sinusoidal shear wave generated by the external vibrator. Shear wave signals from different sub-volumes were aligned in temporal direction to correct time delays from sequential pulse-echo events, followed by 3-D speed reconstruction using a 3-D local frequency estimation algorithm. The technique was validated on liver fibrosis phantoms with different stiffness, showing good correlation (r = 0.99, p < 0.001) with values measured from a state-of-the-art SWE system (GE LOGIQ E9). The phantoms with different stiffnesses can be well-differentiated regardless of the external vibrator position, indicating the feasibility of the 3-D SWE with regard to different shear wave propagation scenarios. Finally, shear wave speed calculated by the 3-D method correlated well with magnetic resonance elastography performed on human liver (r = 0.93, p = 0.02), demonstrating the in vivo feasibility. The proposed technique relies on low volume rate imaging and can be implemented on the widely available clinical ultrasound scanners, facilitating its clinical translation to improve liver fibrosis evaluation.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    39
    References
    1
    Citations
    NaN
    KQI
    []