| 摘要 |
In a method and device for correcting distortion in MRI, k-space data are acquired in a number of data readout directions, the data are converted into a number of images, and a corresponding pixel shift map is generated for each image. The geometric distortion in the corresponding image is corrected according to the pixel shift map, and then all geometric distortion-corrected images are combined. Since movement distortion normally exists in the data readout direction, collecting the k-space data from a number of data readout directions can effectively correct movement distortion. Moreover, correcting the geometric distortion for the images converted from data acquired in a number of data readout directions according to the pixel shift map can reduce the geometric distortion of the final image generated from combination of images. The method and device correct not only movement distortion of MRI images, but also geometric distortion of MRI images.
|
