Last week, the Better AI Health project was featured at the V Congreso Interdisciplinar en Genética Humana in Granada, one of the key European events in genetics and precision medicine, gathering 700+ experts across clinical, genomic, and AI domains .
Within Use Case 2, our team presented a poster on a federated infrastructure for the multimodal analysis of Inherited Retinal Dystrophies, combining clinical, genomic, and imaging data without moving sensitive patient data.
The contribution:
- was accepted for physical poster presentation
- and shortlisted as a candidate for an award
The work demonstrates how federated learning can:
- enable cross-center collaboration while preserving data sovereignty
- integrate heterogeneous healthcare data (clinical, genomic, imaging)
- improve diagnostic pathways in rare diseases
As shown in the results, this approach already proves its potential, reaching up to ~0.95 accuracy in federated training scenarios while keeping data decentralized .
A special mention should be made of Gema García García, leader of Use Case 2, who received the Young Investigator Award from the Spanish Association of Human Genetics during the congress.
