TurtleSeg is an interactive segmentation tool originally designed for 3D medical images. TurtleSeg is developed at the Medical Image Analysis Lab at Simon Fraser University and the Biomedical Signal and Image Computing Laboratory at the University of British Columbia.

Accurate and automatic 3D medical image segmentation remains an elusive goal and manual intervention is often unavoidable. TurtleSeg implements techniques that allow the user to provide intuitive yet minimal interaction for guiding the 3D segmentation process.

A typical workflow involves having the user load a 3D image and then use 2D Livewire to manually contour a sparse number of different slices. The full 3D segmentation can then be built automatically using the user-provided information. TurtleSeg is named after the Turtlemap 3D Livewire algorithm [1] implemented in TurtleSeg for constructing the 3D segmentation. The algorithm employs the concept of a "Turtlemap" in order to automatically produce a dense set of parallel segmentation contours from a sparse set of user provided contours.

TurtleSeg additionally implements a feature called Spotlight, an automated system for automatically directing the user towards the next best place for them to provide input. With Spotlight, there is very little need for the user to understand the underlying algorithm used.

TurtleSeg features:

• Intuitive and easy to use interface (Getting Started Video)
• Spotlight automated assistance for guiding users towards better segmentations
• Natural 3D navigation with support for oblique slice planes (See Views video tutorial)
• Undo/Redo support
• Support for many popular 3D image formats (list)
• Export the segmentation as an image mask or surface mesh
• Full contour editing support makes fixing mistakes easy
• Support for multi-label segmentation (See Multiple Object Segmentation video tutorial)
• Extensive documentation in the form of a manual, FAQ, and video tutorials