Motion detection In-Vivo by multi-channel ultrawideband radar

Abstract

Remote sensing of body and inner organ motion by ultra-wideband radar is a promising alternative to electrocardiogram (ECG) based gating of several diagnostic imaging and image guided therapy modalities, e.g. cardiac magnetic resonance imaging (MRI) and high energy particle therapy. In this paper, we present a noncontact approach for the detection of cardiac mechanics by means of multi-channel UWB radar sensors. To extract the required position information from UWB radar for cardiac MRI the decomposition of respiratory and cardiac motions is performed by applying a blind-source separation method based on several time-delayed second-order correlation matrices to decompose the UWB signals and extract movements of cardiac and respiratory landmarks. We compared the time evolution of these landmarks to the signals of simultaneously measured ECG's demonstrating the strong correlation between cardiac UWB and ECG signals. Furthermore, we implemented a fast 1D MR imaging method in a 3 tesla MR scanner to extract true inner organ landmark movements in real time and present these comparisons with simultaneously acquired multi-channel UWB data.