A novel approach for Doppler blood flow measurement

A new approach to frequency estimation for the velocity estimation in Doppler ultrasound blood flow analysis is presented. The basis of the approach is an adaptive sinusoid-tracking algorithm which is effective in extracting nonstationary signals from within noise and estimating their time-varying parameters, such as the frequency, over time. The preliminary studies conducted using simulated signals show the potential of this approach in estimating Doppler frequency shifts under noisy conditions. A qualitative comparison with the short-time Fourier transform (STFT) is presented to show the advantages of the proposed technique over the STFT. The proposed approach offers advantages over conventional time-frequency analysis techniques in terms of high time-frequency resolution and high noise immunity.