Derive an atlas region from a label — label_to

The rule every atlas pipeline here uses to fill the region column: strip the hemisphere affix, turn brackets, hyphens, underscores and slashes into spaces, lower-case the result and squeeze runs of whitespace.

Usage

label_to_region(label_name, remove_hemi = TRUE, normalize = TRUE)

Arguments

label_name: Character vector of labels.
remove_hemi: Strip hemisphere affixes (default TRUE).
normalize: Lower-case and convert separators to spaces (default TRUE).

Value

A character vector of region names.

Details

Exported because atlas build scripts need to reproduce it exactly. A script that adds a name column keyed on region, for instance, has to key on what the pipeline will actually derive; re-implementing the rule by hand is how a key silently stops matching (an underscore handled but a hyphen not, and one region joins to NA).

The affixes recognised are Left/Right, left/right, lh/rh and L/R, as a prefix or a suffix, separated by -, _, . or a space. A label is never stripped to nothing, so a structure genuinely named "left" keeps its name.

Examples

label_to_region("Left-Thalamus")
#> [1] "thalamus"
label_to_region("Central_Lateral-Lateral_Posterior_Left")
#> [1] "central lateral lateral posterior"
label_to_region(c("Pu_Left", "SNc_PBP_VTA_Right"))
#> [1] "pu"          "snc pbp vta"