Using NiiVue in R Shiny

NiiVue is an open-source JavaScript library for WebGL2-based medical image viewing. This R package allows you to use NiiVue from R Shiny.

Examples

Tractography (TCK, TRK, TRX, VTK)

Demo for NiiVue's Tractography [code]

Time Series Demo

Partial demo for NiiVue's Time Series (toggle thumbnail doesn't work, but the other features do) [code]

Quickstart

1. Install the devtools library

install.packages('devtools')

2. Download this repo

git clone https://github.com/AnthonyAndroulakis/niivue.R

3. Install niivue R library

cd niivue.R/niivue
devtools::install()

4. Use it (for more examples, see the examples folder)

suppressWarnings(library(niivue))
library(shiny)

ui <- fluidPage(
  fluidRow(
    tags$div(style = "width: 100%; height: 600px;", NiivueWidget$new()$plot)
  )
)

server <- function(input, output, session) {
  volumeList <- list(
    list(url = "https://niivue.github.io/niivue/images/mni152.nii.gz",
         colormap = "gray", visible=TRUE, opacity=1)
  )

  nv <- NiivueWidget$new()
  nv$loadVolumes(volumeList)
}

shinyApp(ui, server)

Docs

• you need a shiny server to run niivue.R
• usage mirrors that of the niivue library (camelcase)
• nv$opts and nv$graph are reactiveValues variables

Limitations

• any large data arrays (ex: nv.volumes[0].img) are not sent back to R
• nv$thumbnailVisible(TRUE) doesn't work as intended
• only the onLocationChange niivue event handler is implemented
• no extensive documentation (just refer to https://niivue.github.io/niivue/devdocs/index.html)