R/make.score.img.R

In JMMaronge/medals: Performs Memory Efficient Decomposition for Analysis of Local neighborhood moments for Segmentation

#' A function to create score images for each subject for MEDALS.
#'
#' This function allows you to create the score images for each PC in the MEDALS pipeline. These will be used as predictors in a model to predict lesion areas. This should be the 3rd step in the MEDALS pipeline.
#' @param path.img.list A list of sublists, where each sublist is a list of paths to the files for each imaging modality (ex. T1w, T2w, DWI, etc.) for a particular subject. This has to contain at least one sublist
#' @param path.mask.list A list of paths to the brain mask for each subject. This should be in the same order as the sublists for path.img.list
#' @param loads The loadings matrix to rotate original data by.
#' @param which.scores A vector of the score images desired for each subject. Ex. 1:8.
#' @param verbose Print diagnostic Messages
#' @export
#' @importFrom neurobase check_nifti remake_img
make.score.img<-function(path.img.list,
                         path.mask.list,
                         loads= medals::def.loads,
                         which.scores=1:8,
                         verbose = TRUE){

  score.imgs<-vector(mode = "list",length = length(path.img.list))
  for(i in 1:length(path.img.list)){
    if (verbose) {
      message(paste0("Starting subject ",i))
    }
    x_i <- get.img.moment.dat(
      path.img.list[[i]],
      path.mask.list[[i]],
      mpower = dim(loads)[1]/27/length(path.img.list[[1]]))
    mask <- check_nifti(path.mask.list[[i]])

    if (verbose) {
      message("Rotating Neighborhoods")
    }
    x_i <- x_i %*% loads
    for (l in which.scores) {
      if (verbose) {
        message("Making Score Images")
      }
      score = x_i[, l]
      img <- remake_img(vec = score,
                        img = mask,
                        mask = mask)
      rm(list = "score"); gc(); gc();
      img[ mask == 0 ] = NA
      score.imgs[[i]][[l]] <- img
      rm(list = "img"); gc(); gc();
    }
  }
  return(score.imgs)

}