Denoising of quadrature ultrasound Doppler signal from bi-directional flow based on matching pursuit

摘要

Denoising of Doppler signal is a preliminary and important step in medical ultrasound imaging. To denoise quadrature Doppler signal from bi-directional. flow, we propose a novel method based on matching pursuit in this paper. The proposed method is an iterative decomposition algorithm which decomposes the original Doppler signal into a linear expansion of atoms in a time-frequency dictionary. The time-frequency dictionary is similar to Fourier transform domain and the atoms are similar to orthogonal bases in Fourier transform. In each step of the iteration, the atom which gives the largest inner product with the analyzed signal is selected from the dictionary, and the contribution of this atom is subtracted from the Doppler signal. This process is repeated on the residue until the SNR reaches the maximum. The linear expansion of the selected atoms is the denoised signal. Simulations were conducted on a simulation model with a sampling rate of 12.8 kHz. When the original SNRs are 0 dB, 2 dB, 4 dB, 6 dB, 8 dB, 10 dB, the proposed method can improve the SNR for 7.9 dB, 7.8 dB, 7.5 dB, 7.3 dB, 7.05 dB, 6.8 dB respectively, reduce the root mean square error ( RMSE) of the mean frequency waveform to 0.0441 kHz, 0.0303 kHz, 0.0245 kHz, 0.0215 kHz, 0.0161 kHz, 0.0125 kHz respectively, and suppress the RMSE of the spectral width waveform to 0.1774 kHz, 0.0591 kHz, 0.0486 kHz, 0.0170 kHz, 0.0145 kHz, 0.0117 kHz respectively. Preliminary in vivo evaluation was also carried out on a healthy 33-year-old male using B-K medical A/S 3535 ultrasound scanner, and the results showed that the proposed method can effectively enhance the Doppler spectrogram. (C) 2008 Elsevier B.V. All rights reserved.