2017 Winter Project Week/SegWithSubtractionAndModel

• 
• 

  Projects List

Key Investigators

• Andrew Beers, MGH
• Jayashree Kalpathy-Cramer, MGH
• Alexis G (Kitware)
• Csaba Pinter

Project Description

Objective	Approach and Plan	Progress and Next Steps
In spite of recent advances in automatic segmentation technologies, manual segmentation is still often needed to either correct errors in the automatic segmentation or to achieve higher standard of segmentation accuracy. However, manual segmentation will become increasingly difficult as scan resolution and thus the number of slices to annotate increases. We are developing a module that takes advantage of subtraction maps and automatically-formed 3D ROIs to make accurate and efficient annotations in high resolution spaces.	Confer with others to create convenient 3D visualizations of user-created ROIs, fix UI/UX bugs, integrate with the existing Editor and Segmentations Editor, and package and release as a functioning module.	Bugs were worked out of the program so that it is now fully usable, but the UI and usage changes were put to the side in favor of other projects this week. A great impediment for me in developing python Slicer modules is learning how to translate QT UI items into workable Python code. I haven't been able to find a good guide or examples as to how it applies to the Slicer module interface, and often have to feel my through via experimentation. I would love to be linked to a guide if it exists, but if not I would love to help write one! A lot of time has been spent creating a toolbar button with an icon, for example, and it would be nice to have some existing copy-and-paste code to speed that process along.

Background and References

https://www.researchgate.net/publication/307942551_nD_morphological_contour_interpolation (just in case that helps - Bea)