Recent Changes - Search:

Homepage

This is the old Camino website, please visit our new page to download the code and get the latest documentation.









UCL MIG Home

UCL CS Home

UCL Home

edit SideBar

Man /

Pointset2scheme

Content-type: text/html Manpage of pointset2scheme

pointset2scheme

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

 

NAME

pointset2scheme - Creates scheme files from point sets.

 

SYNOPSIS

pointset2scheme [options]


   

DESCRIPTION

This program reads gradient directions and produces a scheme file. It will read files in the text format found in camino/PointSets/Elec???.txt:


 <N>
 <point 1 x>
 <point 1 y>
 <point 1 z>
 ...
 <point N z>

Alternatively, the program can read files in the format:


 <N>
 <point 1 x> <point 1 y> <point 1 z>
 ...
 <point N x> <point N y> <point N z>

with or without the number of points specified at the top of the file. Gradient directions will be normalized to unit length if non-zero.

The program will also read FSL gradient directions in the format


 <point 1 x> <point 2 x> ... <point N x>
 <point 1 x> <point 2 y> ... <point N y>
 <point 1 x> <point 2 y> ... <point N z>

but it does not read the FSL bvals file. To convert FSL bvecs / bvals file to Camino format, see fsl2scheme(1). This program permits you to define new schemes using gradient directions from FSL.

Normally, directions will have unit magnitude. If this is not the case, you need to determine whether your scanner incorporates the gradient direction magnitude into account when calculating the diffusion gradients. See the -usegradmod option.

The program produces a scheme file appropriate to the supplied imaging parameters.

 

SCHEME FORMATS

Additional details concerning imaging schemes may be found under SCHEME FILES in camino(1). This section is a quick guide to specifying scheme files using pointset2scheme. This program assumes that measurements either apply zero diffusion weighting, or a fixed non-zero diffusion weighting. If this is not the case, it may be easier to write the scheme by hand.

For all schemes created using this program, "zero" measurements, ie those with zero diffusion weighting, should have zero gradient directions.

The simplest scheme format is BVECTOR (formerly called version "2"). This has the form


 VERSION: BVECTOR
 <x_1> <y_1> <z_1> <b_1>
 ...
 <x_N> <y_N> <z_N> <b_N>

where [x_i y_i z_i] is a gradient direction and b_i the corresponding b-value.

To create this scheme file, you need only specify the b-value with the -bvalue option. All non-zero gradient directions will have the specified b-value, and zero gradients will have a zero b-value.

The STEJSKALTANNER scheme format is a fully-specified Stejskal-Tanner sequence (formerly version "1"). Its file format is


 VERSION: STEJSKALTANNER
 <x_1> <y_1> <z_1> <|G|_1> <DELTA_1> <delta_1> <TE_1>
 ...
 <x_N> <y_N> <z_N> <|G|_N> <DELTA_N> <delta_N> <TE_N>

where |G| is the gradient strength, DELTA is the pulse separation, delta is the pulse duration, and TE is the echo time for the measurement. To create a scheme file in this format, use the -dwparams option and pass (in order) <|G|> <DELTA> <delta> <TE_1>. All non-zero measurements will share these parameters. Zero measurements will have zero |G|, DELTA, and delta, but the same TE.

The TRSE scheme format is a fully-specified twice-refocused spin-echo sequence (formerly version "3"). Its file format is


 VERSION: TRSE
 <x_1> <y_1> <z_1> 
 ...
 <x_N> <y_N> <z_N> <|G|>, <del1>, <t_del1>, <del2>, <t_del2>, <del3>, <t_del4>, <TE>

where t_del is the onset and del the duration of a pulse, and there are four pulses per measurement. Details of this scheme may be found in Reese et al, Magn Reson Med, 49(1), 177-82, 2003. This scheme is created by specifying -dwparams followed by (in order) <|G|> <del1> <t_del1> <del2> <t_del2> <del3> <t_del4> <TE>. Zero measurements will have everything but TE set to zero.

The scheme version "0" has been deprecated; it is no longer possible to create scheme files in that format. Existing scheme files can still be read by Camino programs that take the -schemefile option.

 

INCLUDING ZERO MEASUREMENTS

Zero measurements may be placed anywhere in the scheme. A zero measurement is defined as one where the gradient direction is [0 0 0]. If any direction in the input point set is zero, then all diffusion-weighting parameters are also zero for that measurement. Additional zero measurements may be prepended onto the scheme with the -addzeromeas option.

 

COMMENTS

Comments in scheme files are any line beginning with #.

 

UNITS FOR IMAGING PARAMETERS

The units of the measurements in the scheme file determine the units of the output. By default, Camino prefers to use SI: meters, seconds, Tesla. Other units are permitted, but this will affect the output of Camino programs, and may require you to enter additional options. For the purposes of diffusion tensor reconstruction, non-standard units will scale the diffusion tensors in the output. Thus if you use a BVECTOR scheme with b-values specified in s / mm^2, your diffusivities will be in units of mm^2 / s. Other programs may work with non-standard units if additional options are specified. Check the relevant man page before using non-SI units with camino programs.

 

EXAMPLES

Create a scheme file with 30 measurements at b= 1000 s / mm^2 = 1E9 s / m^2, with 3 unweighted measurements followed by the 30 diffusion-weighted measurements. Gradient directions are chosen from the Camino point sets, which are optimised by electrostatic energy minimization (Jansons and Alexander, Inverse Problems, 19:1031-1046, 2003). Many pre-computed point sets are stored in the camino/PointSets directory.


  pointset2scheme -bvalue 1E9 -addzeromeas 3 < PointSets/Elec030.txt > M3_N30_b1000.scheme

Same thing, only this time we use a file points.txt that already contains the vector [0 0 0] for the zero measurements. Note that the zero measurements may appear anywhere in the file, the b-value for these measurements will be set to zero.


  pointset2scheme -bvalue 1E9 -inputfile points.txt > M3_N30_b1000.scheme

Create a scheme where all measurements are repeated to improve SNR.


  pointset2scheme -bvalue 1E9 -inputfile points.txt -numscans 2 > M3_N30_b1000_S2.scheme

Create the same thing, but acquire the repeat measurements sequentially, rather than in a block, so the measurements go [direction 0, direction 0, direction 1, direction 1,...,direction 32] rather than [direction 0, direction 1,..,direction 32, direction 0, direction 1,..,direction 32]


  pointset2scheme -bvalue 1E9 -inputfile points.txt -numscans 2 -interleave > M3_N30_b1000_S2_interleave.scheme

Use a Camino point set, this time for 60 directions, and specify Stejskal-Tanner sequence parameters.


  
  pointset2scheme -version STEJSKALTANNER -dwparams 0.022 0.04 0.025 0.072 -addzeromeas 6 -inputfile PointSets/Elec060.txt > M6_N60_ST.scheme

 

OPTIONS

-version <VERSION>
Version of scheme file to produce. The default is BVECTOR. Options are BVECTOR, STEJSKALTANNER, TRSE.

-bvalue <b>
B-value for all measurements. Applies only to BVECTOR schemes. This is an alias for -dwparams b.

-dwparams <param1 param2 ... paramN>
The diffusion-weighting parameters for each non-zero measurement. The required parameters for each scheme version are


 Version                Parameters


 BVECTOR                b
 STEJSKALTANNER         |g| DELTA delta TE
 TRSE                   |g| del1 t_del1 del2 t_del2 del3 t_del4 TE

-addzeromeas <number>
Prepend number of zero measurements to the scheme. Zero measurements are added before  any repeats are encoded, so these measurements are also repeated.

-numscans <number>
Number of scans, which is the total number of times that the input point set (including any prepended zero measurements) should appear in the scheme file. The default is 1.

-interleave
Interleaves the scans. Useful only if the number of scans is greater than 1. If the input point set contains gradient directions [0, 1, 2,...,N], then -numscans 2 produces output in order [0, 1,...,N, 0, 1,...,N]. -numscans 2 -interleave gives [0, 0, 1, 1,...,N, N].

-flipx
Negate the x component of all the vectors in the input.

-flipy
Negate the y component of all the vectors in the input.

-flipz
Negate the z component of all the vectors in the input.

-usegradmod
This uses the gradient direction magnitude to scale b. Some scanners will factor the gradient magnitude into the b-value. If the gradient directions have magnitude |g| = sqrt(g_x^2 + g_y^2 + g_z^2), and the nominal b-value is beta, then the actual b-value is beta * |g|^2. If this option is used, the gradient directions will still be normalized to unity in the output, and the b-value will be scaled appropriately. It is only used for BVECTOR schemes.

 

AUTHORS

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

 

SEE ALSO

camino(1), fsl2scheme(1)

 

BUGS


 

Index

NAME
SYNOPSIS
DESCRIPTION
SCHEME FORMATS
INCLUDING ZERO MEASUREMENTS
COMMENTS
UNITS FOR IMAGING PARAMETERS
EXAMPLES
OPTIONS
AUTHORS
SEE ALSO
BUGS

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

Edit - History - Print - Recent Changes - Search
Page last modified on October 26, 2009, at 02:59 PM