Medical imaging
SiderUNet
Active-learning MRI segmentation with a reproducible 3D nnU-Net pipeline.
Project detailsWasim Aftab, PhD
I develop reproducible computational methods and research software for complex biomedical data—from multi-omics and spatial biology to medical imaging and knowledge-grounded AI.
Approach
My work spans transcriptomics, single-cell and spatial omics, proteomics and multimodal MRI. At LMU Munich and LMU Klinikum, I have developed statistical methods, reproducible workflows and usable scientific applications, including ImShot for spatial proteomics, WeiseEule for biomedical evidence retrieval, and current methods for cell-type attribution and MRI segmentation.
I enjoy working across the full process: shaping biological questions with collaborators, designing rigorous analyses, building maintainable software for Linux and HPC environments, and translating computational results into clear biological conclusions.
Selected work
Medical imaging
Active-learning MRI segmentation with a reproducible 3D nnU-Net pipeline.
Project detailsSingle-cell & proteomics
Replicate-aware cell-type attribution of bulk proteomic signals.
Project detailsSpatial proteomics
Published open-source software for in-situ protein identification.
Project detailsBiomedical information retrieval
Published evidence retrieval for knowledge-grounded biomedical question answering.
Project details