Content-type: text/html Manpage of shfit

shfit

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

NAME

shfit - Fits the spherical harmonic series to the log measurements.

SYNOPSIS

shfit [options]

DESCRIPTION

Fits the spherical harmonic series to the apparent diffusion coefficient profile. The program computes d = -1/b log(A(q)/A(0)) for each measurements and fits the series to d as a function of q/|q|. The output of the operation in each voxel is the array [exitCode, log A^tar(0), c00, c20, Re(c21), Im(c21), Re(c22), Im(c22), c40, Re(c41), Im(c41), exitCode indicates the success of the operation and should be zero if everything worked OK; A^tar(0) is the best fit MRI signal at b=0.

Fitting the spherical harmonic series is not particularly useful on its own. More useful it to use the spherical harmonic series for voxel classification, see voxelclassify(1). The anisotropy of the spherical harmonic fit, which sfanis(1) can compute, is a useful alternative to the fractional anisotropy, fa(1), which takes into account departures from Gaussian displacements.

EXAMPLES

Fit the fourth-order spherical harmonic series to some synthetic data with two dominant directions:

datasynth -testfunc 3 -voxels 1 -snr 16 -schemefile A.scheme | shfit -schemefile A.scheme -order 4 > test_SH.Bdouble

Compute the spherical harmonic anisotropy:

datasynth -testfunc 3 -voxels 1 -snr 16 -schemefile A.scheme | shfit -schemefile A.scheme -order 4 | sfanis -inputmodel sh -order 4 | double2txt

OPTIONS

Input/output data and simulations options are the same as modelfit(1).

Options concerned with the spherical harmonic fitting:
-order <Maximum series order>: Specifies the highest order of the series to fit.

BUGS

AUTHOR

Danny Alexander <camino@cs.ucl.ac.uk>

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

Shfit