Experimental investigation of transverse velocity estimation using cross-correlation

A method for blood flow estimation has been demonstrated by clinically relevant phantom and in-vivo measurements. A relative standard deviation of 13.1 % and a bias of -2.5 % was obtained with an experimental ultrasound scanner for a purely transverse flow. This precision is within the limits for current normal color flow mapping systems that cannot estimate the transverse flow. A major advantage is the low decorrelation as a result of focusing along the flow, which contributes to the low standard deviation at other angles. Preliminary in-vivo data showed that the approach can be used for large beam-to-flow angles in the clinic. Additional measurements and further development of our experimental ultrasound system is likely to improve these results.