Bjoern MenzeOpen OpportunitiesBackground: Topology is vital in medical image segmentation, emphasizing anatomically correct structures & removing incorrect ones. Previous works [1-3] explored how to enforce topological constraints, however, are applied only at training. Recent diffusion-based models [4, 5] offer a novel way to enforce topological constraints during inference.
MSc Thesis: First, you will develop a diffusion model for segmentation. Next, you will be devising a novel way to integrate topological constraints in the diffusion model. Importantly, we aim to publish the results of this work with you at a high-impact conference or journal. - Artificial Intelligence and Signal and Image Processing, Electrical and Electronic Engineering, Interdisciplinary Engineering, Mathematical Sciences
- Master Thesis
| In this master's thesis project, we are looking for a candidate to apply machine learning techniques to correct and predict signals of incomplete CT perfusion imaging for ischemic stroke. We hope to use machine learning techniques to de-noise and correct for the truncation in CT perfusion signals. In particular, we aim to infer the true attenuation curve after the truncation time-point. - Artificial Intelligence and Signal and Image Processing, Central Nervous System, Radiology and Organ Imaging
- Master Thesis
| This is a clinical image registration and visualization project that tries to map a zoomed-in CT view with a zoomed-out MR modality. The CT view can see very detailed blood vessels and bones, while the MR view sees the soft brain tissues but without vessels. The clinicians want to register them together automatically, as they are currently aligning the two views by hand manually and takes them a lot of time. The outcome of this project is an automated, fast, and accurate image co-registration software that can be deployed in the hospital to improve clinical care. - Information, Computing and Communication Sciences, Medical and Health Sciences
- Internship, Master Thesis, Semester Project
|
|