Sovereign 8s.is mirror of standardgalactic/aegis
- Jupyter Notebook 97.5%
- JavaScript 2.1%
- HTML 0.4%
| Filename | Latest commit message | Latest commit date |
|---|---|---|
| .github/workflows | ||
| _layouts | ||
| docs | ||
| examples | ||
| node_modules | ||
| notebooks | ||
| scripts | ||
| templates | ||
| AEGIS_WEBSITE_PLAN.md | ||
| CITATION.cff | ||
| LICENSE | ||
| package-lock.json | ||
| package.json | ||
| README.md | ||
AEGIS: AI/ML Evaluation and Governance Infrastructure for Safety
An Operational Infrastructure for Post-Market Governance of Adaptive Medical AI Under US and EU Regulatory Frameworks
AEGIS provides a structured, regulatory-aligned infrastructure for governing continuously learning medical AI systems through iterative retraining cycles. It achieves continuous safety oversight through three integrated modules:
- Dataset Assimilation and Retraining Module (DARM),
- Model Monitoring Module (MMM), and
- Conditional Decision Module (CDM).
Quick start
git clone https://github.com/ki-smile/aegis.git
cd aegis
# Open the Jupyter notebook example
jupyter notebook notebooks/AEGIS_Sepsis_Physionet_standalone.ipynb
Links
| Resource | Link |
|---|---|
| Documentation | ki-smile.github.io/aegis |
| Implementation guide | Guide |
| Interactive forms | Forms |
| Sepsis example | Sepsis notebook |
| Brain tumour example | Brain tumour notebook |
Citation
@article{afdideh2026aegis,
title = {AEGIS: An Operational Infrastructure for Post-Market
Governance of Adaptive Medical AI Under US and EU Regulations},
author = {Afdideh, Fardin and Astaraki, Mehdi and Seoane, Fernando
and Abtahi, Farhad},
year = {2026},
journal = {},
doi = {10.XXXX/PLACEHOLDER},
url = {https://ki-smile.github.io/aegis}
}
License
© 2026 SMAILE (Stockholm Medical Artificial Intelligence and Learning Environments), Karolinska Institutet. All rights reserved. See LICENSE for details.