Method employing point source to determine motion induced errors in mr

Abstract

A method is provided for determining errors in MR imaging which result from translational motion of an object. In accordance with the method, an MR point source is rigidly joined to the object in selected spatial relationship, and for movement in unison therewith. An MR system is operated to acquire an overall k-space signal which represents an image of the object and of the point source collectively, the overall k-space signal being contaminated by phase errors which result from the motion. A k-space data set which represents an image of the point source alone, and which remains contaminated by the phase errors, is filtered or separated out from the overall k-space signal. The MR system is operated in selected association with the point source to acquire a reference k-space data set, which represents an image of the point source alone but which is unaffected by the phase errors resulting from the motion. The reference k-space data set is then selectively compared with the filtered k-space data set, to determine the phase errors. Alternatively, the reference k-space data set can be replaced by a plurality of k-space alignment lines in parallel relationship to the phase-encoding axis.