'''Index'''
<<TableOfContents>>

= Name =
mri_ca_train - create a GCA from MNI and xfm files

= Synopsis =
mri_ca_train [options] <subject 1> <subject 2> ... <output gca fname>

= Arguments =
== Positional Arguments ==
|| <subject 1> || subject 1 ||
|| <subject 2> || subject 2 ||
|| <output file> || output file ||

== Required Flagged Arguments ==
|| -parc_dir || segmentation directory (path relative to $subject/mri||

== Optional Flagged Arguments ==
|| -spacing  ||spacing of classifiers in canonical space||
|| -gradient ||use intensity gradient as input to classifier||
|| -mask volname ||use volname as a mask ||
|| -node_spacing ||spacing of classifiers in canonical space i.e. memory you devote to intensity distribution||
|| -prior_spacing ||spacing of class priors in canonical space i.e. memory you devote to prior probabity distributions||
|| -input name || specifying training data (path relative to $subject/mri). can specify multiple inputs.  If not specified, "orig" is used.  Using orig is sufficient for an atlas created with MPRAGES, but for other data types (i.e. multi-echo flash) this volume should be choosen more carefully, see example 2. ||
|| -T1 || input volume (i.e. nu, norm.mgz etc) ||
|| -flash || specifies the atlas as either single-echo or multi-echo ||
|| -xform || reference transform ||

= Outputs =
|| <output file> || gca file - gaussian class array file, in this instance essentially a probabilistic map ||


= Examples =
== Example 1 ==
{{{
mri_ca_train -prior_spacing 2 \
  -node_spacing 8 \
  -mask brain \
  -parc_dir esegatlast \
  -xform manual_registration.xfm \
  -T1 nu subject1 single_one.gca
}}}
== Example 2 ==
{{{
mri_ca_train -node_spacing 4 \
  -prior_spacing 2 \
  -parc_dir esegatlast \
  -xform talairach_one.m3d \
  -input flash30_avg.mgz \
  -input flash5_avg.mgz \
  $SUBJECTS multi_one.gca
}}}

= See Also =
[[SubcorticalSegmentation]], [[mri_em_register]], [[mri_ca_normalize]]

= References =
  * [[https://surfer.nmr.mgh.harvard.edu/ftp/articles/fischl02-labeling.pdf|Whole Brain Segmentation: Automated Labeling of Neuroanatomical Structures in the Human Brain]], Fischl et al., (2002). Neuron, 33:341-355.

= Reporting Bugs =
Report bugs to <analysis-bugs@nmr.mgh.harvard.edu>

= Author/s =
BruceFischl