A Matrix Pencil estimator with adaptive rank selection: Application to In Vivo flow studies

Abstract

Recently, we showed that the rank-two Matrix Pencil method can be used to obtain low-biased flow estimates from in vitro color flow data without clutter filtering. This article addresses how the Matrix Pencil framework can be modified to perform in vivo flow estimation through adaptive selection of the eigen-structure rank. In our modified approach, the nominal rank is defined as the minimum eigen-structure rank that yields principal frequency estimates with a spread greater than a prescribed bandwidth. To examine its performance, the rank-adaptive Matrix Pencil method was applied to in vivo M-mode data and color flow data obtained from the human common carotid artery. It was found that the rank-adaptive Matrix Pencil was able to give a more consistent visualization of the carotid blood flow than the rank-two Matrix Pencil estimator. © 2006 IEEE.