Application of parametric spectral estimation to medical ultrasound and magnetic resonance spectroscopy

Abstract

for Devices and Radiological Health, FDA, Rockville, MD 20857.2. Radiology Dept., Georgetown University Medical Center, Washington, D. C.3. Radiology Dept., University of Kansas, Kansas City, KA.ABSTRACTThis paper addresses variances of estimates of power spectral densities of radio—frequency (RF) signalsgenerated with ultrasound and magnetic resonance spectroscopy. The spectral estimation methods studiedinvolved autoregressive (AR) and moving average (MA) models. With experimental ultrasonic data, thepower spectral density estimate obtained using the MA model exhibited an appreciable reduction in van-ance compared with the squared modulus of the FFT. With magnetic resonance spectroscopic data, theAR spectral estimate was comparable to, but not significantly better than, the squared modulus of theFFT.