While artificial intelligence (AI) has shown promising results in predicting heart failure (HF) outcomes, existing models remain largely proof-of-concept, with limited real-world application—especially within the Spanish population.

TrustAI-ES aims to bridge this gap by developing privacy-preserving AI solutions based on federated learning (FL) and image synthesis. These technologies will unlock previously inaccessible cardiac imaging data, enabling their secure re-use and the local optimization of AI models trained on large, independent datasets to enhance heart failure risk assessment.

Through federated learning, Spanish institutions will collaborate effectively to train AI models without sharing sensitive patient information. In parallel, cardiac image synthesis techniques will strategically expand the available training data, particularly for under-represented groups in Spain, such as women and individuals of non-European ethnic backgrounds. This is essential for ensuring algorithmic fairness and reducing bias in AI models.
Ultimately, TrustAI-ES seeks to create trustworthy AI tools that can be seamlessly integrated into routine clinical practice within the Spanish healthcare system, promoting inclusive, equitable, and reliable healthcare for all members of Spain’s diverse society.