FENITH
Federated Learning in Dynamic and Heterogeneous Environments for European Digital Healthcare: From Theoretical Framework to Implementation in Italian Hospitals
PhD Thesis — Fabio Liberti · ORCID · Google Scholar
Overview
This repository serves as the central hub for the PhD research project on Federated Learning (FL) applied to the European Health Data Space (EHDS). It aggregates all research artifacts — experimental platforms, domain-specific applications, governance frameworks, foundational studies, published papers, and thesis documents — into a single, structured meta-repository.
Research Questions
| |
Question |
Cluster |
| RQ1 |
Technological Framework — How to design FL architectures capable of handling heterogeneity of hospital nodes? |
Platforms |
| RQ2 |
Multidimensional Governance — Which governance models balance innovation, compliance, and ethics? |
Governance |
| RQ3 |
Practical Adoption — What are the barriers and enablers for FL adoption in Italian hospitals? |
Applications |
| RQ4 |
EHDS Interoperability — How to integrate FL with HL7/FHIR, OMOP/OHDSI standards? |
Platforms · Governance |
At a Glance
| |
|
| FL Algorithms |
38 unique across 3 platforms |
| Datasets |
35 total — benchmark, clinical imaging, healthcare |
| Publications |
10 papers — 1 published (21 citations), 3 presented, 2 accepted, 4 submitted |
| Submodules |
16 repositories (14 public + 2 private) |
| Frameworks |
TensorFlow, PyTorch, Flower, Flask, React |
Repository Clusters
| Repository |
Description |
Stack |
| flopbg |
FL platform — Universitas Mercatorum & OPBG |
Python, TensorFlow, React |
| BLEKFL2 |
FL platform — BTH & Universitas Mercatorum |
Python, PyTorch, Flower |
| FL-EHDS-FLICS2026 |
FL + EHDS governance framework |
Python |
Applications
| Repository |
Description |
Domain |
| Questionnaire_FL |
FL adoption questionnaire — Italian hospitals |
Healthcare Adoption |
| CIDE |
OMOP/FHIR business models for EHDS |
Digital Health |
| CIDE2 |
XAI deepfake detection in telemedicine |
XAI |
| FedHR5.0 |
FL for HR Management in Industry 5.0 |
HR |
| CRISTAIN2025 |
FA-FedAvg for law enforcement |
Criminal Justice |
Governance
| Repository |
Description |
Focus |
| AI-DIGOSA |
Norms, ethics, and innovation tensions |
Ethics & Regulation |
| icsis2026 |
FL for territorial healthcare planning |
Health Policy |
| ICID2026 |
Three-layered FL architecture for EHDS |
Architecture |
Foundations
| Repository |
Description |
Type |
| DHFLPL2 |
Foundational paper — MDPI 2024 (21 citations) |
Seminal |
| DHFLPL |
Original published paper repo (v1) |
Legacy |
| Heterogeneous_FL |
Educational materials on heterogeneous FL |
Educational |
Publications
| Code |
Title |
Venue |
Year |
Status |
| P-M |
Federated Learning in Dynamic and Heterogeneous Environments |
MDPI Applied Sciences · DOI |
2024 |
Published |
| P-IS |
FedHR5.0: Privacy-Preserving HR Management in Industry 5.0 |
ISM 2025 |
2025 |
Accepted |
| P-CR |
FA-FedAvg: Forensic-Aware Federated Averaging |
CRISTAIN 2025 |
2025 |
Accepted |
| P-C1 |
Business Models for European Digital Health Research Networks |
CIDE 2025 |
2025 |
Presented |
| P-C2 |
Explainable FL for Secure Telemedicine |
CIDE 2025 |
2025 |
Presented |
| P-IT |
AI Distribuita e Governance Sanitaria |
ITAIS 2025 |
2025 |
Presented |
| FLICS |
FL + EHDS Governance Framework |
FLICS 2026 — IEEE |
2026 |
Submitted |
| ICSIS |
FL as Policy Data Infrastructure |
ICSIS 2026 |
2026 |
Submitted |
| ICID |
Three-Layered Reference Architecture for FL within EHDS |
ICID 2026 — Springer |
2026 |
Submitted |
| IFKAD |
FedHR5.0: FL for Knowledge Asset Dynamics |
IFKAD 2026 |
2026 |
Submitted |
Getting Started
git clone --recurse-submodules https://github.com/FabioLiberti/FENITH2.git
git submodule update --remote --merge
Citation
If you use this work, please cite:
Liberti, F.; Berardi, D.; Martini, B. Federated Learning in Dynamic and Heterogeneous Environments. Applied Sciences 2024, 14(18), 8490. DOI: 10.3390/app14188490
View full README on GitHub · www.fenith.org · License: CC BY-NC-SA 4.0