Created by Leo Seeger, Markus Müller, Ramona Bendias from TU Berlin.
We performed cerebral aneurysm detection and analysis. The data used in the experiments was provided by TU Berlin and the same as provided in the CADA Challenge. This work can be used for aneurysm detection, segmentation and rupture risk estimation.
- templates (Python): Refactored Notebooks that are reusable for a specific task (e.g. model training, prediction, data exploration). Notebooks should be used in the following order: train, create_preds, postprocessing_evaluation. For Unet & PointNet we have different files.
- utils: Utility functions that are used across multiple notebooks/scripts. Installable via pip from project folder:
pip install -e ./utils. - models: The folder contains the final models.