Segmentation

This demo presents the new ImageViz extension to algorithmically segment each vertebra of a Spine.

Two different inputs are used: The original dataset (a stack of images), and a set of markers that we created in a second volume. In most cases, spines are similar for same size subjects, which allows us to load a set of predefined markers and adjust their position manually or automatically.

When the markers are well positioned, we use the original volume data set and the marker data set as inputs to the "watershed" operator. The output of this operator is a new "label" volume data set with a unique integer label value assigned to all the voxels that belong to each of the vertebra.

FILES:

• Medical/Segmentation/medicalSegmentation/medicalSegmentation.cxx

SEE ALSO

SoVolumeData, SoDataRange, SoTransferFunction, SoVolumeRenderingQuality, SoVolumeRender, SoVRImageDataReader, SoImageDataAdapter, SoGradientOperatorProcessing3d, SoMarkerBasedWatershedProcessing

SCREENSHOT: