Thesis topic

Cross-Domain Translation of Angiography: From Stenosis to Healthy Vessel Appearance

Unsupervised translation from diseased to healthy vessel appearance.

Cross-Domain Translation of Angiography: From Stenosis to Healthy Vessel Appearance

Cross-Domain Translation of Angiography: From Stenosis to Healthy Vessel Appearance

Overview

  • Optimal stent choice is essential for optimal long-term therapy outcomes.
  • Knowing the original size and shape of a diseased vessel can support the optimal stent choice.
  • Develop an unsupervised learning framework for data-domain transfer from diseased to healthy vessels, for example with a CycleGAN.
  • Work with coronary and/or iliac angiography data.

Skills

  • Python programming experience with Keras or PyTorch, scikit-learn, pandas, and NumPy
  • Deep learning knowledge, especially CNNs and GANs
  • Image processing knowledge
  • Basic medical or anatomy knowledge is a plus

Tasks

  • Literature research
  • Data preparation and preprocessing pipeline implementation
  • Probabilistic segmentation and/or detection model implementation, training, and evaluation

Contact

Daniel Wulff

d.wulff@uni-rostock.de