Content-type: text/html Manpage of wdtfit

wdtfit

Section: User Commands (1)
Index Return to Main Contents

NAME

wdtfit - Fits the diffusion tensor to diffusion MRI measurements.

SYNOPSIS

wdtfit <data file> <scheme file> [noise var map file] [options]

DESCRIPTION

Reads diffusion MRI data, acquired using the acquisition scheme detailed in the scheme file, and fits the diffusion tensor to the log of the data using a weighted linear least squares regression. For help with scheme files, please see the section "scheme files" in camino(1).

The data file stores the diffusion MRI data in voxel order with the measurements stored as big-endian four-byte floats and ordered as in the scheme file. See modelfit(1) and camino(1) for the format of the data file and scheme file. The program fits the diffusion tensor to each voxel and outputs the results, in voxel order and as big-endian eight-byte doubles, to the standard output. The program outputs eight values in each voxel: [exit code, ln(S(0)), D_xx, D_xy, D_xz, D_yy, D_yz, D_zz]. An exit code of zero indicates no problems. For a list of other exit codes, see modelfit(1). The entry S(0) is an estimate of the signal at q=0.

The weighted fitting is an iterative process; the weights are initialized to the predicted DW signal given the tensor from the ordinary least squares inversion (inversion 1, or the default for dtfit). The weights for the next iteration are the predicted signal from the previous estimate of D. This follows the procedure of Basser and Le Bihan, J Magn Reson B 40:247-254 (1994) and Salvador et al, Human Brain Mapping 24:144-155 (2005).

If any of the data is bad, ie, log(S_i) cannot be computed for any i, then that measurement is excluded from the fitting procedure.

The noise variance map is a map of the estimated variance of the noise in the voxel, assuming that the noise can be approximated by a zero-mean Gaussian distribution (reasonable for a signal to noise ratio > 5). If the weighted fit cannot be calculated for any reason, the output is the unweighted inversion.

The noise variance sigma^2 is the sum of the squares of the weighted residuals normalized by the number of degrees of freedom in the voxel. If there are N diffusion-weighted measurements and none are bad, the degrees of freedom is (N - 7), since there are 7 parameters in the fitting (6 for the DT, plus S_0).

This program is a simplified wrapper to modelfit. The command

wdtfit SubjectA.Bfloat A.scheme

is equivalent to

modelfit -inputfile SubjectA.Bfloat -schemefile A.scheme -inversion 7

The command

dtfit SubjectA.Bfloat A.scheme sigmaSq.Bdouble

is equivalent to

modelfit -inputfile SubjectA.Bfloat -schemefile A.scheme -inversion 7 -noisemap sigmaSq.Bdouble

To read data from stdin, do:

cat <data file> | wdtfit - A.scheme

The handedness of the coordinate system used by the scanner may not agree with that used within Camino. If this happens, the anisotropy and trace of tensors will be correct but the orientations will be incorrect. This problem is often only noticed during tractography. A simple way to check for this is to fit the diffusion tensors and visualize the principal directions with the dtpdview program. If the principal directions appear to be rotated by 180 degrees about the X, Y or Z directions, then the likely cause is that the gradient directions do not agree. This can be remedied by negating the relevant entries in the scheme file.

EXAMPLES

Fit the diffusion tensor to the data in SubjectA.Bfloat using weighted linear regression to the log measurements:

wdtfit SubjectA.Bfloat A.scheme > DiffTensorA.Bdouble

Equivalently:

cat SubjectA.Bfloat | wdtfit - A.scheme > DiffTensorA.Bdouble

AUTHORS

Philip Cook <camino@cs.ucl.ac.uk>

BUGS

This document was created by man2html, using the manual pages.
Time: 02:07:11 GMT, December 04, 2017

Wdtfit