| 摘要 |
<p>A fully sampled data set is generated and stored in a fully sampled memory (42). An undersampled data set corresponding to a portion of the same field of view as the fully sampled data set is generated and stored in an undersampled data memory (62). The fully sampled data set includes a frequency encoded data line corresponding to each of the phase encode gradient steps or angles required to span the selected field of view with a selected resolution. The undersampled data set includes data lines with only a fraction of the phase encoding steps or angles. The fully sampled data set is Fourier transformed (44) into a fully sampled image representation (50) and stored in a fully sampled image memory (46). The undersampled data is Fourier transformed (44) into an undersampled image representation (66) and stored in an undersampled image memory (64). The undersampled image from the views of the undersampled data set between boundaries (58a, 58b) includes a representation of the tissue (52a, 52b, 54) within these boundaries as well as a representation from tissue (52c, 52d) outside of the boundaries superimposed thereon. Portions (72, 74) of the fully sampled image representing tissue outside of the undersampled image are translated and combined (86) and subtracted (88) from the undersampled image representation to generate a corrected undersampled image (90) for display on a video monitor (48). In this manner, a portion of fully sampled image representation that is free of roll-over artifacts is utilized to correct an undersampled, incomplete image representation of the same region that includes the roll-over artifacts.</p> |
