| 摘要 |
<p>A method for generating magnetic resonance (MR) images or MR spectra of at least one partial area of a moving object with at least one motion sequence that is repeated during consecutive motion states is proposed. In a learning measurement, a monitor signal of the repeating motion sequence is recorded, and MR test data of the partial area of the object are recorded under known measurement conditions, wherein the MR test data is associated with the motion states of the motion sequence. In an evaluation step, the MR test data of the motion states of the motion sequence are compared to each other with respect to at least one parameter, and the variation of the at least one parameter over the motion sequence is determined. The instantaneous motion state of the object in the motion sequence is constantly determined in a target measurement during recording of the MR images or MR spectra of the partial area via a monitor signal, and the associated variation of the at least one parameter is constantly compensated for by corresponding variation of the measurement conditions. The quality of MR measurements for recording moving objects is thereby improved, in particular, when retroactive calculation of disturbances in the measured MR data is not possible.</p> |
