Basics

In ggcolonviz: A 'ggplot2' Extension for Visualizing the Colon

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(ggcolonviz)
library(ggplot2)

Basic Plotting

patient_data <- data.frame(
  x = 0,
  y = 0,
  part = c(
    "Cecum",
    "Ileum",
    "Ascending",
    "Transverse",
    "Descending",
    "Sigmoid",
    "Rectum"
  )
)

patient_data |>
  ggplot(aes(x = x, y = y, part = part)) +
  geom_colon(
    color = "black",
    linewidth = 0.5,
    fill = "lightblue",
    show.legend = FALSE
  ) +
  geom_colon_label() +
  coord_fixed() +
  theme_void()

Layout Variants

The SESCD variant is designed for visualizing the Simple Endoscopic Score for Crohn's Disease (SES-CD). It uses a specific layout scheme and labeling to represent the different segments of the colon according to the SES-CD scoring system.

patient_data <- data.frame(
  x = 0,
  y = 0,
  part = c(
    "Ileum",
    "Right",
    "Left",
    "Transverse",
    "Rectum"
  )
)

patient_data |>
  ggplot(aes(x = x, y = y, part = part)) +
  geom_colon(
    color = "black",
    linewidth = 0.5,
    fill = "lightblue",
    variant = "sescd"
  ) +
  geom_colon_label(variant = "sescd") +
  coord_fixed() +
  theme_void()

Custom Labels

You can also provide custom labels for each segment of the colon. This allows you to display specific information or annotations directly on the plot.

patient_data <- data.frame(
  x = 0,
  y = 0,
  part = c(
    "Ileum",
    "Right",
    "Left",
    "Transverse",
    "Rectum"
  ),
  label = c(
    "Ileum",
    "Right Colon",
    "Left Colon",
    "Transverse Colon",
    "Rectum"
  )
)

patient_data |>
  ggplot(aes(x = x, y = y, part = part)) +
  geom_colon(
    color = "black",
    linewidth = 0.5,
    fill = "lightblue",
    variant = "sescd"
  ) +
  geom_colon_label(variant = "sescd", aes(label = label)) +
  coord_fixed() +
  theme_void()

Faceting

Faceting by Subject

Faceting is supported, allowing you to create multiple plots based on a categorical variable. For example:

# Define the parts
parts <- c(
  "Cecum",
  "Ileum",
  "Ascending",
  "Transverse",
  "Descending",
  "Sigmoid",
  "Rectum"
)

# Create subjects with discrete stages
discrete_data <- dplyr::bind_rows(
  data.frame(
    subject = "Subject A: Localized",
    x = 0,
    y = 0,
    part = parts,
    stage = factor(c(0, 0, 0, 1, 3, 2, 2))
  ), # 0=Normal, 1=Mild, 2=Mod, 3=Severe

  data.frame(
    subject = "Subject B: Diffuse",
    x = 0,
    y = 0,
    part = parts,
    stage = factor(c(1, 1, 1, 2, 2, 3, 3))
  ),

  data.frame(
    subject = "Subject C: Remission",
    x = 0,
    y = 0,
    part = parts,
    stage = factor(c(0, 0, 0, 0, 0, 1, 1))
  )
)

ggplot(
  discrete_data, 
  aes(x = x, y = y, part = part, fill = stage)) +
  # 1. Draw the colon segments
  geom_colon(color = "white", linewidth = 0.5) +

  # 2. Add the labels (nudge/center logic still applies)
  # Map color to 'stage' too if you want text to change with background
  geom_colon_label(
    aes(angle = after_stat(angle)),
    size = 2,
    fontface = "bold",
    color = "white"
  ) +

  # 3. Facet by subject
  facet_wrap(~subject) +

  # 4. DISCRETE VIRIDIS
  # 'plasma' or 'viridis' work well for discrete categories
  scale_fill_viridis_d(
    option = "plasma",
    name = "Disease Stage",
    labels = c(
      "0: Normal",
      "1: Mild",
      "2: Moderate",
      "3: Severe"
    )
  ) +
  coord_fixed() +
  theme_void() +
  theme(
    legend.position = "top"
  )

Faceting by Time

You can also create time series plots to show how the condition of the colon changes over time for a patient.

# Longitudinal data for Subject 101
longitudinal_data <- dplyr::bind_rows(
  # Baseline: Severe inflammation in the lower colon
  data.frame(
    visit = "Week 0",
    x = 0,
    y = 0,
    part = parts,
    stage = factor(c(0, 0, 0, 1, 3, 3, 3), levels = 0:3)
  ),

  # Week 4: Partial response, inflammation moving down
  data.frame(
    visit = "Week 04",
    x = 0,
    y = 0,
    part = parts,
    stage = factor(c(0, 0, 0, 0, 2, 2, 3), levels = 0:3)
  ),

  # Week 8: Significant improvement
  data.frame(
    visit = "Week 08",
    x = 0,
    y = 0,
    part = parts,
    stage = factor(c(0, 0, 0, 0, 0, 1, 2), levels = 0:3)
  ),

  # Week 12: Clinical Remission
  data.frame(
    visit = "Week 12",
    x = 0,
    y = 0,
    part = parts,
    stage = factor(c(0, 0, 0, 0, 0, 0, 1), levels = 0:3)
  )
)

ggplot(longitudinal_data, aes(x = x, y = y, part = part, fill = stage)) +
  # 1. Custom Colon Geometry
  geom_colon(color = "white", linewidth = 0.4) +

  # 2. Centered Labels (nudge logic applies to every facet)
  geom_colon_label(
    aes(angle = after_stat(angle)),
    size = 2.2,
    fontface = "bold",
    color = "white"
  ) +

  # 3. Facet by Visit (Time Point)
  facet_wrap(~visit, nrow = 1) +

  # 4. Discrete Viridis Scale
  scale_fill_viridis_d(
    option = "magma", # 'magma' highlights 'hot' spots well
    name = "Disease Stage",
    labels = c("0: Normal", "1: Mild", "2: Moderate", "3: Severe"),
    drop = FALSE # Keep all stages in the legend even if not present in a facet
  ) +

  coord_fixed() +
  theme_minimal() +
  theme(
    panel.grid = element_blank(),
    axis.text = element_blank(),
    axis.title = element_blank(),
    strip.text = element_text(size = 12, face = "bold"),
    legend.position = "bottom"
  ) +
  labs(
    title = "Subject 101: Colon Disease Progression Over 12 Weeks")

Custom Blueprints

ggcolonviz ships with two built-in layout variants ("standard" and "sescd"), but you can register your own blueprint for any anatomical scheme you need.

How a blueprint works

A blueprint is a plain data frame with three columns:

| Column | Type | Meaning | |--------|------|---------| | dx | numeric | Relative x-coordinate of each polygon vertex | | dy | numeric | Relative y-coordinate of each polygon vertex | | part | character | Name of the segment this vertex belongs to |

Each unique part value must have at least 3 rows (the minimum for a closed polygon). The coordinates are offsets — when you supply x = 5, y = 2 in your plot data, every vertex for that row is shifted by (5, 2).

Listing available variants

list_colon_variants()

Registering a custom variant

Use register_colon_variant() to add your blueprint to the package registry. The variant is then available by name in geom_colon(), geom_colon_label(), colon_sf(), and colon_heatmap_grid().

# A minimal two-segment blueprint (e.g. a simplified proximal / distal split)
simple_blueprint <- data.frame(
  dx   = c(-2, 0, 0, -2,    0,  2,  2,  0),
  dy   = c( 0, 0, 3,  3,    0,  0,  3,  3),
  part = c(rep("Proximal", 4), rep("Distal", 4))
)

register_colon_variant("simple", simple_blueprint)

# Confirm it is registered
list_colon_variants()

Using a custom variant in a plot

Once registered, the variant works identically to the built-in ones:

data.frame(x = 0, y = 0, part = c("Proximal", "Distal")) |>
  ggplot(aes(x = x, y = y, part = part)) +
  geom_colon(
    variant  = "simple",
    color    = "black",
    linewidth = 0.5,
    fill     = "lightblue"
  ) +
  geom_colon_label(variant = "simple") +
  coord_fixed() +
  theme_void()

Overwriting a variant

Calling register_colon_variant() with an existing name replaces it. This lets you iterate on your blueprint interactively:

# Refine the blueprint and re-register
simple_blueprint_v2 <- simple_blueprint
simple_blueprint_v2$dy <- simple_blueprint_v2$dy * 1.5   # taller segments

register_colon_variant("simple", simple_blueprint_v2)

Validation

register_colon_variant() validates the input and stops with a clear message if:

• name is not a single non-empty string
• data is not a data frame
• Any of the required columns (dx, dy, part) are missing
• dx or dy are not numeric
• Any segment has fewer than 3 vertex rows

# Missing the 'dy' column — will error informatively
bad_bp <- data.frame(dx = c(0, 1, 1), part = c("A", "A", "A"))
register_colon_variant("bad", bad_bp)

ggcolonviz documentation built on Oct. 15, 2026, 5:09 p.m.