Differences

This shows you the differences between two versions of the page.

Link to this comparison view

reference:ft_headmovement [2018/08/23 14:43] (current)
Line 1: Line 1:
 +=====  FT_HEADMOVEMENT =====
 +
 +Note that this reference documentation is identical to the help that is displayed in MATLAB when you type "help ft_headmovement"​.
 +
 +<​html><​pre>​
 +  <a href=/​reference/​ft_headmovement><​font color=green>​FT_HEADMOVEMENT</​font></​a>​ creates a raw data structure, or cell-array of datastructures
 +  containing the HLC-coil data, which have a grad structure that has the
 +  head position information incorporated.
 + 
 +  Use as
 +    data = ft_headmovement(cfg)
 + 
 +  where the configuration should contain
 +    cfg.dataset ​     = string with the filename
 +    cfg.method ​      = '​updatesens'​ (default), '​cluster',​ '​avgoverrpt',​
 +                         '​pertrial_cluster',​ '​pertrial'​
 + 
 +  optional arguments are
 +    cfg.trl ​         = empty (default), or Nx3 matrix with the trial 
 +                         ​definition,​ can be empty.see <a href=/​reference/​ft_definetrial><​font color=green>​FT_DEFINETRIAL</​font></​a>​. If
 +                         ​defined empty, the whole recording is used
 +    cfg.numclusters ​ = number of segments with constant headposition in
 +                         which to split the data (default = 10). This
 +                         ​argument is used in some of the methods only (see
 +                         ​below),​ and is used in a kmeans clustering scheme.
 + 
 +  If cfg.method = '​updatesens',​ the grad in the single output structure has
 +  a specification of the coils expanded as per the centroids of the position
 +  clusters. The balancing matrix is s a weighted concatenation of the
 +  original tra-matrix. This method requires cfg.numclusters to be specified
 + 
 +  If cfg.method = '​avgoverrpt',​ the grad in the single output structure has
 +  a specification of the coils according to the average head position
 +  across the specified samples.
 + 
 +  If cfg.method = '​cluster',​ the cell-array of output structures represent
 +  the epochs in which the head was considered to be positioned close to the
 +  corresponding kmeans-cluster'​s centroid. The corresponding grad-structure
 +  is specified according to this cluster'​s centroid. This method requires
 +  cfg.numclusters to be specified.
 + 
 +  If cfg.method = '​pertrial',​ the cell-array of output structures contains
 +  single trials, each trial with a trial-specific grad structure. Note that
 +  this is extremely memory inefficient with large numbers of trials, and
 +  probably an overkill.
 + 
 +  If cfg.method = '​pertrial_clusters',​ the cell-array of output structures
 +  contains sets of trials where the trial-specific head position was
 +  considered to be positioned close to the corresponding kmeans-cluster'​s
 +  centroid. The corresponding grad-structure is specified accordin to the
 +  cluster'​s centroid. This method requires cfg.numclusters to be specified.
 + 
 +  The updatesens method and related methods are described by Stolk et al., Online and
 +  offline tools for head movement compensation in MEG. NeuroImage, 2012.
 + 
 +  See also <a href=/​reference/​ft_regressconfound><​font color=green>​FT_REGRESSCONFOUND</​font></​a>​ <a href=/​reference/​ft_realtime_headlocalizer><​font color=green>​FT_REALTIME_HEADLOCALIZER</​font></​a>​
 +</​pre></​html>​