Backend server for PARTEA (Privacy-Aware Time-to-Event Analysis), a federated platform for cross-institutional survival analysis without sharing raw patient data.
PARTEA enables multiple hospitals or research institutes to jointly run survival analyses (Kaplan-Meier, Cox regression, federated survival SVMs) on their combined patient cohorts, while keeping all individual-level data on-premise.
This repository contains the coordination server, which:
- Orchestrates federated computation rounds across participating clients
- Aggregates partial statistics (never raw patient data)
For the user-facing web interface, see federated-partea/webapp.
- Python, Django
- PostgreSQL
- Docker
- Lifelines
Späth et al. (2022). Privacy-aware multi-institutional time-to-event studies. PLOS Digital Health. Link to paper
See LICENSE.