MSc Thesis: transfer learning for segmentation of tubular structures in thoracic CT images
Computed tomography (CT) scans are commonly used in clinic practice to diagnose and monitor diseases of the lung, heart and upper abdomen . Deep learning has seen wide-spread application for the segmentation of organs in CT images [2,3]. However, its use for segmentation of complex, tubular structures, such as the bronchial proximal airways or cardiac vasculature, has been inhibited by a lack of high-quality ground truth labels . In this research project, the prospective student will investigate the use of unsupervised learning and transfer learning to segment complex structures in chest CT images without having to rely on large amounts of ground truth annotations.
- Enthusiasm for deep learning and biomedical imaging.
- Advanced knowledge of machine learning and computer vision. Ideally, prior work experience using deep learning for image processing.
- Excellent programming skills in Python as well as PyTorch or Tensorflow.
- Full time commitment towards the completion of your Master’s project.
What we offer:
- An exciting research project with many possibilities to bring in your own ideas.
- Close supervision and access to state-of-the-art computer hardware.
- The chance to work in a team of experts in image processing, deep learning, biomedical engineering and medicine.
How to apply:
Send an email to email@example.com and firstname.lastname@example.org with your CV and transcript. We aim to get back to you within a couple of days.
 Grainger, Ronald G., and David J. Allison, eds. Grainger & Allison’s diagnostic radiology: a textbook of medical imaging. Vol. 1. Churchill Livingstone, 1997.
 Dong, Xue, et al. “Automatic multiorgan segmentation in thorax CT images using U‐net‐GAN.” Medical physics 46.5 (2019): 2157-2168.
 Wasserthal, Jakob, et al. “TotalSegmentator: robust segmentation of 104 anatomical structures in CT images.” arXiv preprint arXiv:2208.05868 (2022).
 Willemink, Martin J., et al. “Preparing medical imaging data for machine learning.” Radiology 295.1 (2020): 4-15.