Two-dimensional spectrum estimation for flow with multiple velocities using maximum entropy

摘要

We have developed a high-resolution method to estimate multiple velocities of flows within one sample volume of pulse echo ultrasound for small flows. The method uses maximum entropy method (MEM) to estimate the covariance matrix of a two-dimensional (2D) signal when data length is limited. The estimated covariance is used to calculate high-resolution 2D power spectrum of flows. The simulation and experiment results show that our method provides high-resolution estimations for multiple velocities when data length is limited and small flows are located.