论文部分内容阅读
我们将四种谱分析技术用于脉冲多普勒超声正交信号以比较每种技术在估算血流速度和多普勒谱方面的相对优点。这四种技术是:1)快速傅里叶变换怯;2)最大似然法;3)Burg自回归算法;4)修正的协方差算法到自回归建模。我们对模拟信号和从在体血流系统获得的信号均作了研究。我们确定了每种方法的最佳参数值(例如模型阶数)并研究了信噪比和信号带宽的影响。选择适当的模型阶数可以发现现代
We applied four spectral analysis techniques to pulsed Doppler ultrasound orthogonal signals to compare the relative merits of each technique in estimating blood flow velocity and Doppler spectrum. The four techniques are: 1) fast Fourier transform; 2) maximum likelihood method; 3) Burg autoregressive algorithm; 4) modified covariance algorithm to autoregressive modeling. We studied both the analog signal and the signal obtained from the in vivo blood flow system. We determined the best parameter values (such as model order) for each method and studied the effects of signal-to-noise ratio and signal bandwidth. Choose the appropriate model order can be found in modern times