We are seeking collaboration from:

1.    Data-sharing partners with access to clinical and genomic data on rare hematological diseases.

2.    AI and data science experts to refine and enhance our predictive models.

3.    Medical imaging specialists to integrate imaging data (CT, MRI, PET-CT, histology) into our platform.

4.    Platform developers to standardize and scale the models for testing and industrialization in other areas of medicine.

We have developed a federated learning infrastructure that enables collaboration among multiple stakeholders (hospitals, pharmaceutical companies, researchers) without centralizing or sharing data. This has been successfully deployed and validated in the GenoMed4All Consortium, involving three clinical research centers (IRCCS Humanitas Research Hospital, Universität Leipzig, and the University of Bologna). Results were submitted to the 66th American Society of Hematology Congress 2024.

The project has achieved significant milestones in federated learning, ensuring data privacy while enabling large-scale collaboration for enhanced AI capabilities. Considering this progress, we are currently at Technology Readiness Level (TRL) 3-4. To integrate this technology into innovative solutions, such as a generative AI platform for producing synthetic data, a dedicated initiative is needed to industrialize the technology and engineer the developed components. This would ensure the transition from research to practical, scalable applications, enabling the deployment of advanced AI-driven solutions in real-world healthcare settings.

The results are replicable for other diseases if the technology and the developed modules are engineered into a dedicated platform tailored for specific purposes. By creating a specialized infrastructure, this approach could be extended to address various medical conditions, enabling the deployment of AI-driven solutions across different areas of healthcare with high precision and scalability.