Diplomarbeit / Masterarbeit / Bachelor thesis (DA/MA/BA): Automatic fiducial and implant segmentation using C-arm fluoroscopy
Advisor: Prof. Dr. Nassir Navab
Supervision by: Dr. Pascal Fallavollita
Subsequent 3D reconstruction of implanted seeds within the prostate requires knowledge of their 2D pixel positions as well as the pose of the fluoroscopy image. Pose tracking of a C-arm is a major technical problem that presently does not have a clinically practical solution in the prostate brachytherapy context. Image-based tracking with radiographic fiducial is an inexpensive and practical option. Combinations of beads and parametric curves constrain the optimization during pose recovery, leading to sub-millimetre and sub-degree accuracy in pose tracking. Despite significant efforts, radiographic fiducial based tracking has not gained clinical traction due to a major technical problem: automatic segmentation of the fiducial motifs inevitably fails in the operating room, which necessitates manual intervention that disrupts the entire clinical workflow.
The aim of the project is to eliminate explicit segmentation of the fiducial from the pose tracking process and to automatically segment the implants inside the images for subsequent 3D reconstruction. An automatic level set formulation will be implemented that will lock on both the implant and fiducial features automatically for each fluoroscopy image.
- Interest in medical imaging and computer assisted surgery.
- Basic knowledge of segmentation and computer vision is recommendable.
What do we offer
- Experience different fields of research in computer science: Computer Vision, Visualization, Software Engineering.
- A possibility of solving a still-pending clinical problem.
- Your thesis and degree. (and also submission to prestigious conferences/journals depending on student performance)
If you are interested please contact Dr. Pascal Fallavollita