bold2betasCCA: Convert a bold time series and design matrix to event-wise...

In stnava/RKRNS: ANTsR Decoding

Description

Uses cca and nuisance variables to estimate multivariate betas per event.

Usage

bold2betasCCA(boldmatrix, designmatrix, blockNumb, bl = 12, baseshift = 5,
  mask = NA, sparseness = c(0, 0), multievents = FALSE, polydegree = 10,
  bestvoxnum = 50, nvecs = 5, whichcols = NA, mycoption = 1, its = 10,
  onlyhrf = FALSE)

Arguments

boldmatrix	input raw bold data in time by space matrix
designmatrix	input design matrix - binary/impulse entries for event related design, blocks otherwise
blockNumb	numbers for the rows that should be treated together as runs
bl	basis length for hrf estimation
baseshift	number of time points to ignore post-event onset
polydegree	number of polynomial predictors
nvecs	number of cca predictors to explore e.g. 5

Value

returns a list with relevant output

Author(s)

Avants BB

Examples

# get example image
fn<-paste(path.package("RKRNS"),"/extdata/111157_mocoref_masked.nii.gz",sep="") 
eximg<-antsImageRead(fn,3)
fn<-paste(path.package("RKRNS"),"/extdata/subaal.nii.gz",sep="") 
mask<-antsImageRead(fn,3)
bb<-simulateBOLD(option="henson",eximg=eximg,mask=mask)
mat<-timeseries2matrix( bb$simbold, bb$mask )
runs<-bb$desmat$Run; 
### cca betas
mysp<-c( -0.01, -0.9 )
btsc<-bold2betasCCA( data.matrix(mat)  , bb$desmat[,1:4], blockNumb=runs,
     bl=25, baseshift=0, sparseness=mysp, bestvoxnum=10, mask=mask, 
     polydegree=1, mycoption=1, its=12, onlyhrf=FALSE )
plot(ts(t(btsc$runhrfs)))
# from here see the help for bold2betas

stnava/RKRNS documentation built on May 30, 2019, 7:21 p.m.