A consistent tissue attenuation coefficient estimator using bubble harmonic echoes

The ultrasonic property of soft tissue can be quantified by its attenuation coefficient α. Traditionally the backscattering signal of tissue is used to estimate α. To improve precision, a large number of spatially independent samples of tissue echoes are required for averaging. In this paper, we propose a new estimation method, which makes use of microbubbles to provide temporally independent samples for averaging. It is easier for temporal sampling to maintain ergodicity and provide a large number of independent samples for statistical averaging. A stochastic model for the harmonic signals of an ideal bubble attenuated by tissue is derived based on Kuc's and Miller's works. An estimator of α is then presented. This estimator is consistent and could be biased because of the unknown squarelaw relation between the second and fundamental harmonics for non-ideal bubble oscillation. In experimental works, we design a simplified phantom for demonstrating the performance of the proposed estimator. It is shown that both first and second harmonics can estimate α consistently. However, the interference of the tissue backscattering signal may cause additional estimation error using the first harmonic.