Non-stationary clutter rejection in ultrasound color flow imaging

For ultrasound color flow images with high quality, it is important to suppress the clutter signals originated from tissue. Traditional clutter rejection methods such as FIR, IIR and regression filters are based on the hypothesis that the clutter signal is stationary. While in realistic ultrasonic imaging, there exists considerable acceleration of tissue motion, which makes the clutter signal a non-stationary stochastic process. Under this condition, traditional clutter rejection filters will bring large bias to the flow velocity estimation. In this paper, a new clutter filter scheme was proposed for rejecting non-stationary clutter signal. The scheme first estimated the instantaneous frequency of the clutter using phase modeling method, and then mixed down the demodulated Doppler signals with the estimated frequency. After down mixing, the clutter was attenuated using a traditional high pass filter. Experimental results show that non-stationary clutter signal can be efficiently attenuated using this filtering scheme, especially for low blood flow velocity and high amplitude clutter conditions.