Volume Rendering Component in the upcoming X3D v4

Web3D Consortium member Vicomtech has been working to support volume rendering on the Web platform in any browser. MIRROR4ALL by Vicomtech uses DICOM Slices. This implementation is the construction of the density data from DICOM slices. Try this application, just drag and drop your DICOM images. As a contributor to the X3DOM open source library Vicomtech is helping to transition the X3D standard into the Web HTML5 ecosystem.

Web3D Consortium's X3D Medical Working Group and it's member Vicomtech are collaborating in order to extend the capabilities of the current X3D volume rendering component towards the next iteration of the standard: X3D v4. Vicomtech has been working to support volume rendering in the Web platform without requiring the use of any browser plug-in. 

MIRROR4ALL by Vicomtech - Use your DICOM Slices - drag and drop This implementation is the construction of the density data from DICOM slices, implemented entirely in Javascript and deployed within a browser page. The significance of this is that the user can prepare the visualization from DICOM images on their own computer without the need to send the images to a server for processing. 

When the set of scans is available (3D volume), the original plane direction of each 2D image can be converted into another plane such as coronal, sagital or oblique.  This allows doctors to see the internals of the 3D volume from a perspective that would not be possible if only the base images would be viewed independently. 

The application is built on the X3D standard, an ISO standard for interactive 3D scenegraphs which supports implementations on all platforms: desktop, mobile, VR, and in a WebGL capable web browswer. 

-- One component of the X3D standard is Volume Rendering, enabling the 3D visualization of density data such as obtained in medical scans.
-- The density data for visualization can be obtained by interpolation from a series of DICOM images and incorporated into an X3D file.
-- Open source code for this is available in Python.

Medical scans acquired either from Computed Tomography (CT) or Magnetic Resonance Imaging (MRI) are stored as a series of 2D images along an axis, e.g.: axial plane. This set of images, as a whole, represents the scan of a 3D volume.

On one hand, the ImageTextureAtlas node is a structure that allows to store the set of 2D images that composes the 3D volume in a web compatible manner. It has been used in several demos and examples during a long period of time, since this feature makes possible the ubiquitous visualization of volume rendering content in any modern web browser.

On the other hand, the Multi Planar Reconstruction (MPR) is a visualization tool of general use in the medical domain. It can be employed in medical diagnosis from bone fracture or displacement assessment to pre-operative planning of endovascular treatment of aortic disease.

When the set of scans is available (3D volume), the original plane direction of each 2D image can be converted into another plane such as coronal, sagital or oblique.  This allows doctors to see the internals of the 3D volume from a perspective that would not be possible if only the base images would be viewed independently. 

Based on the feedback from the community and the medical working group, the proposed ImageTextureAtlas and MPRVolumeRenderingStyle nodes are incuded in the next revision of the X3D standard - X3Dv4.

Please visit the X3DOM examples web page https://www.x3dom.org/examples/ to see these extensions, and is available in the upcoming X3DOM release.

Press Release: 
No
Release Date: 
Thu, 2019-10-31