MR compatible sensor for motion artifact corrected reconstruction method

INTRODUCTION: Patient motion, especially respiratory and cardiac motion, leads to artifacts which could damage the quality of Magnetic Resonance Imaging (MRI) and make the resulting diagnosis less accurate. If such motion occurs during MRI acquisitions, it has been shown that, knowing this motion, it is possible to invert the process of artifact production and to reconstruct a motion compensated image [1]. For this purpose, a predictive model of patient motion would be helpful. Several models have already been proposed, based on linear prediction, using navigator echo signals as inputs [2]. GRICS (Generalized Reconstruction by Inversion of Coupled Systems) [3] is a reconstruction strategy for free breathing acquisitions which uses a motion model based on prior knowledge provided by external sensors, such as respiratory belts. Unlike navigators, external sensors are continuously available and do not require any dedicated pulse sequence. Consequently, a new MR compatible external sensor has been developed which aims at measuring the acceleration of a localized region of the body. It gives additional information about patient respiratory motion during MR examination and has been integrated to the GRICS reconstruction as a physiological input to reduce motion artifacts and improve image quality.