Master Thesis Automatic CT/MR image registration

Elekta’s Leksell Gamma Plan supports planning for stereotactic brain radiotherapy delivered with the Leksell Gamma Knife. Accurate treatment depends on mapping structures in CT and MR images to the machine’s coordinate system using markers on a frame attached to the patient’s head. This thesis will investigate how to fully automate the current semi-automatic registration process and whether its accuracy can be improved.

The assignment

• Develop and prototype methods that automatically detect fiducial markers in CT and MR images.
• Develop and prototype fully automatic image registration methods using these markers.
• Define suitable quality assurance metrics, such as mean and maximum errors.
• Assess the robustness and accuracy of the proposed methods against the existing fiducial-based registration in Leksell Gamma Plan using more than 100 varied, representative cases.

Who we are looking for

• Master’s student, preferably with a specialization in image processing or graphics.
• Previous knowledge of Gamma Knife treatment planning is not needed.

What we offer

• Guidance from domain experts on relevant software architecture and clinical workflows.
• Close collaboration with specialists in software development and clinical applications throughout the thesis.
• Supervision by Ethan Palmiere.

Practical details

Location: Elekta’s Stockholm office. Planned start: early 2027. Supervisor contact: Ethan Palmiere, ethan.palmiere@elekta.com.