Estimating ultrasonic parameters of biological tissue by adaptive filtering technique

Diagnostic ultrasonic imaging is widely used as a non-invasive technique for examining the body. However, the resolution of images still does not meet the order of the wavelength, and only a part of the information on tissue characteristics contained in reflected signals is used. In this paper we propose a method for estimating two tissue parameters; i.e., the reflection coefficient and attenuation coefficient from the reflected ultrasound for the purpose of improving the resolution of ultrasonic images and obtaining information on tissue characterization. The reflection coefficient is estimated by a deconvolution technique using a Kalman filter, taking into account the distortion of the propagating pulse due to frequency-dependent attenuation. The attenuation is estimated by adaptive processing based upon the criterion function calculated using estimates of the reflection coefficient. Simulated signals are used to investigate the ability of this method. Additionally, experimental results for reflected signals from a human liver show that the method is applicable to clinical cases.