Shear wave speed estimation by adaptive random sample consensus method

Haoming Lin,Tianfu Wang,Siping Chen
DOI: https://doi.org/10.3233/BME-130832
Abstract:This paper describes a new method for shear wave velocity estimation that is capable of extruding outliers automatically without preset threshold. The proposed method is an adaptive random sample consensus (ARANDSAC) and the metric used here is finding the certain percentage of inliers according to the closest distance criterion. To evaluate the method, the simulation and phantom experiment results were compared using linear regression with all points (LRWAP) and radon sum transform (RS) method. The assessment reveals that the relative biases of mean estimation are 20.00%, 4.67% and 5.33% for LRWAP, ARANDSAC and RS respectively for simulation, 23.53%, 4.08% and 1.08% for phantom experiment. The results suggested that the proposed ARANDSAC algorithm is accurate in shear wave speed estimation.
What problem does this paper attempt to address?