cv

RWTH Aachen University - Ph.D. in Physics

• Thesis topic consists of generative modeling on radiological images.

RWTH Aachen University - M.Sc in Physics

• GPA 1.1/1.0
• Graduate with Distinction (mit Auszeichnung).
• Thesis title is Sequence Optimization for Parameter Quantification in MR Fingerprinting, the grade is 1.0/1.0.
• Master's degree requirements completed while working towards obtaining my Ph.D.

Nankai University - B.Sc in Physics

• GPA 3.6/4.0
• 3^rd out of 60 students.

Uniklinik RWTH Aachen - Research Assistant

Aachen, Germany

• Collaborate with Dr. Daniel Truhn, supervising physician.
• Leading a team working on radiological data synthesis using diffusion models.
• Research on disease progression prediction by latent space exploration on the learned manifold, published in Nature Machine Intelligence, and selected as the cover image of volume 4 issue 11.
• Investigated adversarial robustness on medical and pathological data, through the lens of batch normalization and attention. Both works were published in Nature Communications.
• Research on federated machine learning in medical imaging, work published in Science Advances.

The Alan Turing Institute - Data Study Group

London, UK

• Research topic covers perfusion quantification of sub-lingual micro-circulation.
• Performed unsupervised vessel segmentation on dark field microscopy videos.

Philips Research Hamburg & Uniklinik RWTH Aachen - Master Student

Hamburg & Aachen, Germany.

• Topic is Dictionary-free reconstruction of quantitative MRI.
• 12-month Master thesis project exploring robust relaxation parameter estimation from accelerated MRI measurements, with application to disease quantification in both brain and liver. I applied convex optimization and parallelized the code to scale well. This work was published as conference abstracts in the International Society for Magnetic Resonance in Medicine (ISMRM).

Adversarial representation learning for medical image analysis

CIBSS - Centre for Integrative Biological Signalling Studies, University of Freiburg

Predicting disease progression by using style-based latent extrapolation

Umbrella Symposium

Synthesizing high-resolution medical images using GANs

Radiology department, Uniklinik Köln

Privacy protected federated learning, a generative solution

Philips Research Hamburg

Master in Data Science, Simon Liu

• Thesis topic is Learning Radiological Transferable Features via Cross CT-Xray Contrastive Learning.

Master in Electrical Engineering, Arne Schneuing

• Thesis topic is Reconstruction of Undersampled MRF Data Using Artificial Neural Networks.