Getting started with ggseg

If you work with neuroimaging data, you’ve probably spent time wrestling region-level results into brain plots. ggseg handles that plumbing for you. It stores brain atlas geometries and plots them as ggplot2 layers, so you get brain figures with the same code you’d use for any other ggplot.

library(ggseg)
library(ggplot2)

What’s in a brain atlas

A ggseg_atlas object bundles four things: a name, a type (cortical, subcortical, or tract), region geometry, and an optional colour palette.

dk()
#> 
#> ── dk ggseg atlas ──────────────────────────────────────────────────────────────
#> Type: cortical
#> Regions: 35
#> Hemispheres: left, right
#> Views: inferior, lateral, medial, superior
#> Palette: ✔
#> Rendering: ✔ ggseg
#> ✔ ggseg3d (vertices)
#> ────────────────────────────────────────────────────────────────────────────────
#>    hemi                  region                      label
#> 1  left                bankssts                lh_bankssts
#> 2  left caudalanteriorcingulate lh_caudalanteriorcingulate
#> 3  left     caudalmiddlefrontal     lh_caudalmiddlefrontal
#> 4  left          corpuscallosum          lh_corpuscallosum
#> 5  left                  cuneus                  lh_cuneus
#> 6  left              entorhinal              lh_entorhinal
#> 7  left                fusiform                lh_fusiform
#> 8  left        inferiorparietal        lh_inferiorparietal
#> 9  left        inferiortemporal        lh_inferiortemporal
#> 10 left        isthmuscingulate        lh_isthmuscingulate
#>                              display         lobe
#> 1  banks of superior temporal sulcus     temporal
#> 2          caudal anterior cingulate    cingulate
#> 3              caudal middle frontal      frontal
#> 4                    corpus callosum white matter
#> 5                             cuneus    occipital
#> 6                         entorhinal     temporal
#> 7                           fusiform     temporal
#> 8                  inferior parietal     parietal
#> 9                  inferior temporal     temporal
#> 10                 isthmus cingulate    cingulate
#> ... with 60 more rows

The dk atlas (Desikan-Killiany) ships with the package. Call plot() for a quick look:

plot(dk())

Quick overview of the Desikan-Killiany cortical atlas.

Two helpers pull out the names you’ll need for matching data:

library(ggseg.formats)

sort(unique(atlas_regions(dk())))
#>  [1] "bankssts"                 "caudalanteriorcingulate" 
#>  [3] "caudalmiddlefrontal"      "corpuscallosum"          
#>  [5] "cuneus"                   "entorhinal"              
#>  [7] "frontalpole"              "fusiform"                
#>  [9] "inferiorparietal"         "inferiortemporal"        
#> [11] "insula"                   "isthmuscingulate"        
#> [13] "lateraloccipital"         "lateralorbitofrontal"    
#> [15] "lingual"                  "medialorbitofrontal"     
#> [17] "middletemporal"           "paracentral"             
#> [19] "parahippocampal"          "parsopercularis"         
#> [21] "parsorbitalis"            "parstriangularis"        
#> [23] "pericalcarine"            "postcentral"             
#> [25] "posteriorcingulate"       "precentral"              
#> [27] "precuneus"                "rostralanteriorcingulate"
#> [29] "rostralmiddlefrontal"     "superiorfrontal"         
#> [31] "superiorparietal"         "superiortemporal"        
#> [33] "supramarginal"            "temporalpole"            
#> [35] "transversetemporal"
sort(unique(atlas_labels(dk())))
#>  [1] "lh_bankssts"                 "lh_caudalanteriorcingulate" 
#>  [3] "lh_caudalmiddlefrontal"      "lh_corpuscallosum"          
#>  [5] "lh_cuneus"                   "lh_entorhinal"              
#>  [7] "lh_frontalpole"              "lh_fusiform"                
#>  [9] "lh_inferiorparietal"         "lh_inferiortemporal"        
#> [11] "lh_insula"                   "lh_isthmuscingulate"        
#> [13] "lh_lateraloccipital"         "lh_lateralorbitofrontal"    
#> [15] "lh_lingual"                  "lh_medialorbitofrontal"     
#> [17] "lh_middletemporal"           "lh_paracentral"             
#> [19] "lh_parahippocampal"          "lh_parsopercularis"         
#> [21] "lh_parsorbitalis"            "lh_parstriangularis"        
#> [23] "lh_pericalcarine"            "lh_postcentral"             
#> [25] "lh_posteriorcingulate"       "lh_precentral"              
#> [27] "lh_precuneus"                "lh_rostralanteriorcingulate"
#> [29] "lh_rostralmiddlefrontal"     "lh_superiorfrontal"         
#> [31] "lh_superiorparietal"         "lh_superiortemporal"        
#> [33] "lh_supramarginal"            "lh_temporalpole"            
#> [35] "lh_transversetemporal"       "rh_bankssts"                
#> [37] "rh_caudalanteriorcingulate"  "rh_caudalmiddlefrontal"     
#> [39] "rh_corpuscallosum"           "rh_cuneus"                  
#> [41] "rh_entorhinal"               "rh_frontalpole"             
#> [43] "rh_fusiform"                 "rh_inferiorparietal"        
#> [45] "rh_inferiortemporal"         "rh_insula"                  
#> [47] "rh_isthmuscingulate"         "rh_lateraloccipital"        
#> [49] "rh_lateralorbitofrontal"     "rh_lingual"                 
#> [51] "rh_medialorbitofrontal"      "rh_middletemporal"          
#> [53] "rh_paracentral"              "rh_parahippocampal"         
#> [55] "rh_parsopercularis"          "rh_parsorbitalis"           
#> [57] "rh_parstriangularis"         "rh_pericalcarine"           
#> [59] "rh_postcentral"              "rh_posteriorcingulate"      
#> [61] "rh_precentral"               "rh_precuneus"               
#> [63] "rh_rostralanteriorcingulate" "rh_rostralmiddlefrontal"    
#> [65] "rh_superiorfrontal"          "rh_superiorparietal"        
#> [67] "rh_superiortemporal"         "rh_supramarginal"           
#> [69] "rh_temporalpole"             "rh_transversetemporal"

Your first brain plot

geom_brain() works like any ggplot2 geom. Pass it an atlas and it draws the regions:

ggplot() +
  geom_brain(atlas = dk(), show.legend = FALSE)

A bare atlas: geom_brain() fills by label from the palette.

Arranging views

Brain atlases have multiple views (lateral, medial) and hemispheres. position_brain() controls how they’re laid out, using formula syntax borrowed from facet_grid():

ggplot() +
  geom_brain(
    atlas = dk(),
    position = position_brain(hemi ~ view),
    show.legend = FALSE
  )

Brain views arranged using formula syntax in position_brain().

Or reorder views with a character vector:

ggplot() +
  geom_brain(
    atlas = dk(),
    position = position_brain(c(
      "right lateral",
      "right medial",
      "left lateral",
      "left medial"
    )),
    show.legend = FALSE
  )

Brain views arranged using a character vector to specify order.

See vignette("positioning-views") for the full set of layout options, including grid layouts for subcortical and tract atlases and per-view zoom.

Showing only certain hemispheres or views

ggplot() +
  geom_brain(atlas = dk(), view = "lateral", show.legend = FALSE)

Brain plot showing only lateral views.

ggplot() +
  geom_brain(atlas = dk(), hemi = "left", show.legend = FALSE)

$Brain plot showing only the left hemisphere.$

Brain plot showing only the left hemisphere.

For subcortical atlases, views are slice identifiers. Check what’s available with ggseg.formats::atlas_views():

atlas_views(aseg())
#> [1] "axial_3"   "axial_4"   "axial_5"   "axial_6"   "coronal_1" "coronal_2"
#> [7] "sagittal"

Plotting your own data

Create a data frame with at least one column that matches the atlas (typically region or label). Pass it to geom_brain() through the data argument and it joins to the atlas automatically:

library(dplyr)

some_data <- tibble(
  region = sort(unique(atlas_regions(dk())))[1:4],
  p = sample(seq(0, 0.5, 0.001), 4)
)

ggplot() +
  geom_brain(
    atlas = dk(),
    data = some_data,
    position = position_brain(hemi ~ view),
    aes(fill = p)
  ) +
  scale_fill_viridis_c(option = "cividis", direction = -1) +
  theme_void()

Brain plot coloured by external p-values using a viridis scale.

See vignette("external-data") for more on data preparation and matching.

Faceting

ggplot2 faceting works as you’d expect. Group your data by the faceting variable and geom_brain() replicates the full atlas in each panel:

some_data <- tibble(
  region = rep(sort(unique(atlas_regions(dk())))[1:4], 2),
  p = sample(seq(0, 0.5, 0.001), 8),
  group = c(rep("A", 4), rep("B", 4))
)

ggplot() +
  geom_brain(
    atlas = dk(),
    data = group_by(some_data, group),
    position = position_brain(hemi ~ view),
    aes(fill = p)
  ) +
  facet_wrap(~group)

Brain plots faceted by group, each panel showing the full atlas.

Highlighting specific regions

To colour a few regions with their atlas colours, use two columns: one for the join (region) and one for the fill aesthetic:

data <- data.frame(
  region = sort(unique(atlas_regions(dk())))[1:3],
  reg_col = sort(unique(atlas_labels(dk())))[1:3]
)

palette <- atlas_palette(dk())[1:3]

ggplot() +
  geom_brain(atlas = dk(), data = data, aes(fill = reg_col)) +
  scale_fill_brain_manual(palette)

Highlighting selected regions using the atlas colour palette.

Scales and themes

With neither fill nor colour mapped, geom_brain() applies the atlas palette for you, keyed on label. For custom palettes, use scale_fill_brain_manual().

Built-in themes strip axes and grids for cleaner figures:

ggplot() +
  geom_brain(atlas = dk(), show.legend = FALSE) +
  theme_brain()

Going further

For full control over the data before it reaches ggplot2 — adding region labels, layering other sf geoms, or using geom_sf() directly — see vignette("geom-sf").

Finding more atlases

ggseg ships with dk() (cortical), aseg() (subcortical), suit() (cerebellar flamap), and tracula() (white matter tracts). Many more are available through the ggsegverse r-universe:

install.packages("ggsegYeo2011", repos = "https://ggsegverse.r-universe.dev")