Edit this page

NA-MIC Project Weeks

Back to Projects List

Annotation of Neurosurgical MR and Ultrasound Images with Corresponding Landmarks

Key Investigators

Project Description

Corresponding landmarks between MR and ultrasound images acquired during neurosurgery are valuable for (a) validation of registration algorithms and (b) training supervised registration algorithms (c) initializing a registration. In this project we aim to create a tool that makes the process of finding those landmarks easier.


  1. Objective A. Create a UI that provides new functionality and gathers existing functionality in one place to facilitate landmarking
  2. Objective B. Investigate the rendering infrastructure that would facilitate the adjustment of landmark position in the 3D view of Slicer

Approach and Plan

  1. We use an iterative process for creating the UI - the user(s) give feedback to the developer(s) who then continuously update(s) the UI

Progress and Next Steps


  1. The extension is ready. It can be found here on the main branch. A screenshot can be seen below in Illustrations. For more details refer to the readme.
  2. A lot of bug fixes
  3. More intuitive control of active views
  4. More fine-grained control of viewing options
  5. Automatically join corresponding landmarks with curves to visualise brain shift (also sanity check - the curves should be more or less smooth)

Next Steps

  1. Fulfill all formal requirements for a pull request
  2. Search for bugs/corner cases
  3. Push to the ExtensionIndex

Next Steps (outside the scope of this project week)

  1. Add volume rendering
  2. Automatically detect landmarks (e.g. 3D-SIFT features) and manually choose the best ones


Current state of the extension Screenshot of the current state of the extension

Landmark flow

Landmark flow

Example landmarks


Background and References

  1. Current version of the extension
  2. Mini dataset based on RESECT[1] to use for testing the extension

[1] Xiao, Yiming, et al. “RE troSpective Evaluation of Cerebral Tumors (RESECT): A clinical database of pre‐operative MRI and intra‐operative ultrasound in low‐grade glioma surgeries.” Medical physics 44.7 (2017): 3875-3882.