Iterative reconstruction of the ultrasound attenuation coefficient from backscattered signals

Estimation of the local acoustic attenuation from backscattered signals has several clinical applications but remains an open problem. Most of the proposed solutions relate the observed spectral changes directly to the theoretical predictions. However, these methods make a number of approximations and require correction strategies for acoustic wave phenomena that are not accounted for in the model (e.g., non-linearity). In this study, attenuation was estimated by successively solving the forward wave propagation problem for different attenuation coefficients and by matching the calculated backscattered signals to the observed one. For the forward problem, the effects of attenuation, nonlinear distortion, reflection and scattering were taken into account. The proposed approach was validated on simulated data and data recorded in six tissue mimicking phantoms and was compared to conventional methods. The relative error of the attenuation coefficient remained below 10% for the simulated and phantom data. The...