Medical image computing · geometric deep learning · generative models
I am a Ph.D. student in the Department of Quantitative Biomedicine at the University of Zurich. I work on machine learning methods for understanding and generating complex biomedical structures, particularly vascular anatomy.
Research output
Expand “View research figure” for a preview.
Learning compact tokens for intricate 3D vascular and airway graphs, enabling reconstruction, generation, and completion.
VesselTok pipeline overview.
Sparse representation learning for organ-scale vessel structures with efficient reconstruction and generative modeling.
Original publication PDF; navigate to its method figure.
Discrete graph diffusion with semantic corrections for anatomically plausible vessels and airway structures.
Overview of semantically consistent discrete diffusion.
A two-stage denoising diffusion model for synthesizing vascular graphs, including complex connectivity.
3D vessel graph diffusion overview.
Directly predicting radius-aware 3D vessel graphs from images and merging patch-wise predictions into full-volume networks.
VesselFormer Figure 2 from the published MIDL paper.
Self-supervised medical image learning with masked reconstruction, structure-aware losses and contrastive masked views.
ViT-AE++ Figure 1 (original paper architecture).
Graph neural modeling of brain MR images for predicting inflammatory disease activity.
Figure 1 from the paper: lesion detection, graph processing and self-pruning for disease activity prediction.
Knowledge-guided classification of previously unseen patterns in chest radiographs.
Original publication PDF; navigate to its method figure.
About
Department of Quantitative Biomedicine
Thesis: Natural Language Based 3D Object Localization in Volumetric Grids.
Munich, Germany
Hyderabad, India
TopCoW and TopBrain MICCAI challenges
EXCITE Summer School — Applications of Machine Learning in Medical Domain