Obtain various spatial features of an MR image, which are used in tissue classification.
makeMRIspatial(mask, nnei, sub, bias)A list containing the following components:
a matrix, each row of which giving the neighbors of a
voxel or subvoxel. The number of rows is equal to the number of
(sub)voxels within the mask and the number of columns is the
number of neighbors of each (sub)voxel. For the (sub)voxels on the
boundaries, when one or more of their
neighbors are missing, the missing are represented by the total
number of (sub)voxels within the mask plus 1.
the (sub)voxels within each block are mutually independent given the (sub)voxels in other blocks.
logical; the same as the input sub.
if sub is TRUE, it is a matrix,
with each row giving the eight subvoxels of a voxel;
otherwise it is equal to NULL.
if bias is TRUE, it is a vector
of weights of neighbors of every voxel for bias field correction;
otherwise it is equal to NULL. The default is NULL.
if bias is TRUE, it is a vector
of sum of weights of neighbors for bias field correction, one
element per voxel;
otherwise it is equal to NULL. The default is NULL.
three dimensional array. The voxels with value 1 are inside the mask; with value 0 are outside. We just focus on voxels inside the mask.
the number of neighbors. Right now only 6, 18, and 26 neighbors are supported. For a 3D image, besides defining 6 neighbors in the x, y, and z directions, one can add 12 diagonal neighbors in the x-y, x-z, and y-z planes, and another 8 on the 3D diagonals. This leads to a six neighbor structure, an eighteen neighbor structure, and a twenty-six neighbor structure.
logical; if TRUE, a new mask
which splits each voxel into
eight subvoxels is generated, and then obtain the neighbors and blocks
of subvoxels; otherwise obtain the neighbors and blocks at
the voxel level. The default if FALSE.
logical; if TRUE, the spatial parameters for biased
field correction are calculated. The default if FALSE.
Dai Feng, Dong Liang, and Luke Tierney (2013) An unified Bayesian hierarchical model for MRI tissue classification Statistics in Medicine
Dai Feng (2008) Bayesian Hidden Markov Normal Mixture Models with Application to MRI Tissue Classification Ph. D. Dissertation, The University of Iowa
mask <- array(1, dim=c(2,2,2))
spa <- makeMRIspatial(mask, nnei=6, sub=FALSE)
spa <- makeMRIspatial(mask, nnei=6, sub=TRUE)
spa <- makeMRIspatial(mask, nnei=26, sub=TRUE, bias=TRUE)
Run the code above in your browser using DataLab