ich_predict: Predict ICH Images

In neuroconductor/ichseg: Intracerebral Hemorrhage Segmentation of X-Ray Computed Tomography (CT) Images

Description

This function will take the data.frame of predictors and predict the ICH voxels from the model chosen.

Usage

ich_predict(
  df,
  nim,
  model = c("rf", "logistic", "big_rf"),
  verbose = TRUE,
  native = TRUE,
  native_img = NULL,
  transformlist = NULL,
  interpolator = NULL,
  native_thresh = 0.5,
  shiny = FALSE,
  model_list = NULL,
  smoothed_cutoffs = NULL,
  outfile = NULL,
  ...
)

Arguments

df	data.frame of predictors. If multiplier column does not exist, then ich_candidate_voxels will be called
nim	object of class nifti, from make_predictors
model	model to use for prediction, either the random forest (rf) or logistic
verbose	Print diagnostic output
native	Should native-space predictions be given?
native_img	object of class nifti, which is the dimensions of the native image
transformlist	Transforms list for the transformations back to native space. NOTE: these will be inverted.
interpolator	Interpolator for the transformation back to native space
native_thresh	Threshold for re-thresholding binary mask after interpolation
shiny	Should shiny progress be called?
model_list	list of model objects, used mainly for retraining but only expert use.
smoothed_cutoffs	A list with an element mod.dice.coef, only expert use.
outfile	filename for output file. We write the smoothed, thresholded image. If native = TRUE, then the file will be native space, otherwise in registered space
...	Additional options passsed to ich_preprocess

Value

List of output registered and native space prediction/probability images

See Also

ich_candidate_voxels

neuroconductor/ichseg documentation built on Sept. 29, 2020, 2:31 p.m.