nifti-imaging

Neuroimaging Informatics Technology Initiative (NIfTI) image rendering pipeline for Node.js and browser using the nifti-reader-js library.

Note

This effort is a work-in-progress and should not be used for production or clinical purposes.

Install

Node.js

npm install nifti-imaging

Browser

<script type="text/javascript" src="https://unpkg.com/nifti-imaging"></script>

Build

npm install
npm run build

Features

• Renders single and multi-slice datasets with optional adjustment of window/level and color palette.
• Handles all standard NIfTI datatypes, including integer, float and complex voxel types.
• Decodes gzip-compressed NIfTI files (.nii.gz) transparently.
• Supports 4D time-series datasets.
• Outputs RGBA pixel arrays, suitable for use with HTML5 Canvas, WebGL and WebGPU, or other imaging libraries.
• Provides a common bundle for both Node.js and browser.

Supported Datatypes

• UINT8 (Type code 2)
• INT8 (Type code 256)
• UINT16 (Type code 512)
• INT16 (Type code 4)
• UINT32 (Type code 768)
• INT32 (Type code 8)
• FLOAT32 (Type code 16)
• FLOAT64 (Type code 64)
• COMPLEX64 (Type code 32) — rendered as voxel magnitude
• RGB24 (Type code 128) — true-color

Usage

Basic image rendering

// Import objects in Node.js
const niftiImaging = require('nifti-imaging');
const { NiftiImage } = niftiImaging;

// Import objects in Browser
const { NiftiImage } = window.niftiImaging;

// Create an ArrayBuffer with the contents of the NIfTI byte stream.
const image = new NiftiImage(arrayBuffer);

// Render image.
const renderingResult = image.render();

// Rendered pixels in an RGBA ArrayBuffer.
const renderedPixels = renderingResult.pixels;
// Rendered width.
const width = renderingResult.width;
// Rendered height.
const height = renderingResult.height;

Advanced image rendering

// Import objects in Node.js
const niftiImaging = require('nifti-imaging');
const { NiftiImage, WindowLevel } = niftiImaging;
const { StandardColorPalette } = niftiImaging.constants;

// Import objects in Browser
const { NiftiImage, WindowLevel } = window.niftiImaging;
const { StandardColorPalette } = window.niftiImaging.constants;

// Create an ArrayBuffer with the contents of the NIfTI byte stream.
const image = new NiftiImage(arrayBuffer);

// Create image rendering options.
const renderingOpts = {
  // Optional slice index.
  // If not provided, the first slice is rendered.
  slice: 0,
  // Optional time-point index for 4D datasets.
  // If not provided, the first time-point is used.
  timePoint: 0,
  // Optional user-provided window/level.
  // If not provided, the rendering pipeline calculates it from pixel values.
  windowLevel: new WindowLevel(windowWidth, windowLevel),
  // Optional flag to indicate whether histograms should be calculated.
  // If not provided, the histograms are not calculated.
  calculateHistograms: false,
  // Optional standard color palette.
  // If not provided, the grayscale palette is used.
  colorPalette: StandardColorPalette.Grayscale
};

// Render image.
const renderingResult = image.render(renderingOpts);

// Rendered pixels in an RGBA ArrayBuffer.
const renderedPixels = renderingResult.pixels;
// Rendered slice index.
const slice = renderingResult.slice;
// Rendered time-point index.
const timePoint = renderingResult.timePoint;
// Rendered width.
const width = renderingResult.width;
// Rendered height.
const height = renderingResult.height;
// Window/level used to render the pixels.
const windowLevel = renderingResult.windowLevel;
// Array of calculated per-channel histograms.
// In case calculateHistograms rendering option is false
// histograms should not be present.
const histograms = renderingResult.histograms;

Please check a live example here.

License

nifti-imaging is released under the MIT License.