Application of autoregressive spectral analysis for ultrasoundattenuation estimation: interest in highly attenuating medium

摘要

The authors deal with the application of parametric spectralnanalysis for attenuation estimation on the reflected ultrasound signalnof biological tissues. A second-order autoregressive (AR2) model, whosenparameters are estimated with the Burg algorithm, is used to estimatenthe center frequency on echo signals and its evolution versus depth.nData simulation of independent A-lines backscattered by a homogeneousnmedium of scatterers are generated by a computer model with attenuationnvalues ranging from 1 to 5 dB/cmMHz, an ultrasonic frequency of 5 MHznand different pulse durations. The performance of the estimator isnevaluated for time windows ranging from 5 to 0.3 Μs. The comparisonnis made with the classical short time Fourier analysis using a fastnFourier transform (FFT). It is found that the AR2 model provides anbetter estimation of attenuation than the Fourier technique: thenrelative error of attenuation is below 5% for windows between 0.6 to 2.5nΜs, while the one obtained with the Fourier technique lies between 3nand 16% for the same window sizes. However, the variance of attenuationnestimate is the same with the two techniques. These results offernpromises for determining attenuation in highly attenuating mediumn(material or biological tissue) either because of their structure ornbecause high frequencies are used