Mask ROIs can be used in a number of different contexts — such as for training Deep Learning models, computing Watersheds, and generating vector fields for an in-depth analysis. When applied, masks limit computations to a subset of the original data, which can help reduce processing times and increase accuracy.
You can create a mask ROI one of two ways. You can remove all labeled voxels from a selected region of interest that are outside of a shape (see Creating Mask ROIs by Removing Labeled Voxels from an ROI), or you can add labeled voxels to a selected ROI that are inside a shape (see Creating Mask ROIs by Adding Labeled Voxels to an ROI).
In this case, you will be removing all of the labeled voxels from a selected region of interest that are outside of a shape, such as a box, cylinder, or sphere.
Original ROI and shape (on left) and Mask ROI (on right)

The shape is added to the Data Properties and Settings panel and appears in the workspace.

All labeled voxels that are outside of the shape are removed from the selected ROI.
In this case, you will be adding labeled voxels to a selected region of interest that are within a shape, such as a box, cylinder, or sphere.
Defined shape (on left) and after creating a mask ROI (on right)
The new ROI, which does not contain any labeled voxels, is added to the Data Properties and Settings panel.

The shape is added to the Data Properties and Settings panel and appears in the workspace.


All voxels that are inside the shape are labeled.