Spectroscopy and image texture analysis

Abstract

Acoustic spectroscopy concerns the estimation of acoustic parameters of tissue by means of analysis of the frequency (spectral) contents of ultrasonographic signals (or images). Conventional echographic imaging is based on signals that are processed within the equipment in such a way that the spectral information is distorted. For this reason, the radiofrequency (RF) signals have to be used after linear amplification (i.e., prior to demodulation and logarithmic compression). A further prerequisite is the removal of the influence of beam formation, at transmission and at reception, on the spectral information. This influence becomes evident in echographic images by the depth dependence of the speckle pattern, both in the apparent echo level and in the size of individual speckles (Oosterveld et al. 1985). When all these influences are appropriately accounted for, it becomes possible to estimate the attenuation coefficient and the backscattering coefficient, and their frequency dependence, of parenchymal tissues. A further step is to perform the estimation locally and to construct two-dimensional (2-D) parametric images. The analogy to current developments in magnetic resonance spectroscopy may be evident.