This repository contains the code used for the paper "Towards automated brain aneurysm detection in TOF-MRA: open data, weak labels, and anatomical knowledge "
ADAM challenge submission in MICCAI 2020, from the Kubiac team
In-silico vascular research platform for virtual cardiovascular populations.
Case definitions, validation table, and entry-point scripts for 'Functional and Ethical Modeling of Non-Newtonian Hemodynamics in Aneurysmal and Bifurcated Vessels Using PINNs' (Scientific Culture, Oct 2025). Not a full reproduction — see Status.
Patient-grouped aneurysm rupture classification, generalization across datasets and anatomical locations, and FEM-informed hemodynamic modeling.
Weakly annotate aneurysms in 3D Slicer with two clicks — a sphere from a pair of points, for building detection datasets without segmentation.
Research on cerebral aneurysm rupture classification using vascular geometry, hemodynamics, clinical variables, and deep learning.
Review AI-detected aneurysm candidates in 3D Slicer — verdicts, rejection reasons, sign-off and provenance for clinical validation studies.
End-to-end automated ML pipeline predicting cerebral aneurysm rupture risk in 24 minutes at zero cost, with patient-specific SHAP explanations. AUC 0.834. 1st in Biomedical Engineering Research @ SV, 5th Medicine Research @ CSEF 2026.
To associate your repository with the aneurysm topic, visit your repo's landing page and select "manage topics."