Method of Improved Scatterer Size Estimation without Attenuation Known a Priori

摘要

Backscattered signals contain frequency-dependent information about tissue microstructures, but when using these signals to construct conventional ultrasound B-mode images, the frequency-dependent information will not be utilized. Scatterer size is an important parameter to differentiate diseased tissue from normal. To get the more correct estimation about scatterer size, the more correct assumption about attenuation must be learned. In most cases, the attenuation estimation must be known a priori, and then the scatterer size is computed by a minimal routine. In most case, the attenuation is known a priori. Any errors in the attenuation using the method will produce errors in the resulting scatterer size estimated. Though there are some methods to improve the scatterer size estimation, but most of them also need the attenuation which is known a priori. A new method is proposed in this paper by an analysis of the in vivo backscattered power spectrum, to improve scatterer size estimation. Here the attenuation and the scatterer size were estimated simultaneously, so the attenuation shouldn't be known a priori. The attenuation and scatterer size can be found by searching the attenuation and scatterer values while operating the minimal routine in the new method. Simulations were done to test the new method, the transducer had 8 MHz center frequency, 4 f-number and 5 cm focal length, the attenuation was 0.5 dB/cm/MHz. The result showed that the scatter size (error less than 20%) estimation could be found with large window length (6-12mm), and a better estimation could be found if concerning both total and local attenuation.