Adaptive dynamic quadrature demodulation with autoregressive spectral estimation in ultrasound imaging.

Abstract: In medical ultrasound imaging, the frequency-dependent attenuation causes a downshift of the center frequency of transmitted ultrasound as it propagates through the body. The downshifting results in a considerable loss of signal-to-noise ratio (SNR) after quadrature demodulation (QDM) in which down-mixing and low pass filtering are involved. To overcome the problem, dynamic QDMs have been proposed, in which the change in the center frequency along the axial direction is obtained using autocorrelation-based spectral estimation and compensated in the QDM block. As an alternative, this paper proposes an adaptive dynamic QDM using the 2nd-order autoregressive model. The main advantage over the conventional dynamic QDMs is to use real radio-frequency (RF) data in the spectral estimation, while its counterparts require additional steps to obtain either complex RF signals or complex baseband signals. This allows the proposed method to be used with a minimal modification of signal processing blocks. The performances of the proposed method were evaluated through in vitro and in vivo experiments. The performances were also compared with those of the conventional dynamic QDM. From the experiments, it was learned that the proposed method improved SNR by maximally 7.8dB in the near field compared with the conventional dynamic QDM. In the far field, however, its SNR improvement is similar to its counterpart. This may be explained by the fact that the signal loss mainly results from the amplitude attenuation and the diffraction rather than the frequency downshift in the far field. In addition, the proposed method improved contrast resolution (CR) by at least 6.8%, compared with that of the conventional dynamic QDM. The experimental results demonstrated that the proposed method can be used to improve SNR and CR of ultrasound images in an effective manner. [Copyright &y& Elsevier]