Efficient method for MR image reconstruction using coil sensitivity encoding
SENSitivity Encoding (SENSE) has demonstrated potential for significant scan time reduction using multiple receiver channels. SENSE reconstruction algorithms for non-uniformly sampled data proposed to date require relatively high computational demands. A Projection Onto Convex Sets (POCS)-based SENSE reconstruction method (POCSENSE) has been recently proposed as an efficient reconstruction technique in rectilinear sampling schemes. POCSENSE is an iterative algorithm with a few constraints imposed on the acquired data sets at each iteration. Although POCSENSE can be readily performed on rectilinearly acquired k-space data, it is difficult to apply to non-uniformly acquired k-space data. Iterative Next Neighbor re-Gridding (INNG) algorithm is a recently proposed new reconstruction method for non-uniformly sampled k-space data. The POCSENSE algorithm can be extended to non-rectilinear sampling schemes by using the INNG algorithm. The resulting algorithm (POCSENSINNG) is an efficient SENSE reconstruction algorithm for non-uniformly sampled k-space data, taking into account coil sensitivities.
1. A method, comprising:
acquiring non-rectilinearly sampled k-space data from a set of magnetic resonance (MR) coils having a set of coil sensitivity profiles;
distributing the non-rectilinearly sampled k-space data onto a first resealed matrix having a first scaling factor, where the distributing is performed without using a density compensation function (DCF) and without using a singular value decomposition (SVD) regularization parameter adjustment;
reconstructing subsets of a first MR image from the first resealed matrix by applying a basic Iterative Next Neighbor Gridding (INNG) process to subsets of the first resealed matrix;
selectively zeroing portions of the subsets of the first MR image;
building the first MR image from the subsets of the first MR image;
placing the first MR image in a second rescaled matrix having a second scaling factor;
modifying the non-rectilinearly sampled k-space data into modified k-space data based, at least in part, on the set of coil sensitivity profiles and sensitivity encoding (SENSE);
distributing the modified k-space data onto the second rescaled matrix, where the distributing is performed without using a DCF, the second scaling factor being larger than the first scaling factor;
reconstructing subsets of a second MR image from the second rescaled matrix by applying the basic INNG process to subsets of the second rescaled matrix;
selectively zeroing portions of the subsets of the second MR image;
building the second MR image from the subsets of the second MR image; and
displaying the second MR image.
2. The method of claim 1 , including:
placing the second MR image in a third rescaled matrix having a third scaling factor, the third scaling factor being greater than the second scaling factor;
re-modifying the non-rectilinearly sampled k-space data into re-modified k-space data based, at least in part, on the set of coil sensitivity profiles and sensitivity encoding (SENSE);
distributing the re-modified k-space data onto the third rescaled matrix, where the distributing is performed without using a DCF;
reconstructing subsets of a third MR image from the third rescaled matrix by applying the basic INNG process to subsets of the third rescaled matrix;
selectively zeroing portions of the subsets of the third MR image;
building the third MR image from the subsets of the third MR image; and
displaying the third MR image.
3. The method of claim 2 , including:
placing the third MR image in a fourth rescaled matrix having a fourth scaling factor, the fourth scaling factor being larger than the third scaling factor;
distributing the re-modified k-space data onto the fourth rescaled matrix, where the distributing is performed without using a DCF;
reconstructing subsets of a fourth MR image from the fourth rescaled matrix by applying the basic INNG process to subsets of the fourth rescaled matrix;
selectively zeroing portions of the subsets of the fourth MR image;
building the fourth MR image from the subsets of the fourth MR image; and
displaying the fourth MR image.
4. The method of claim 1 , where distributing the non-rectilinearly sampled k-space data includes determining a location to place a data value by rounding off an original k-space co-ordinate after multiplying by a scale factor.