GMV leads a Consortium that will create a federated network to speed up the use of AI in healthcare systems
GMV, along with 15 other entities participating as a consortium, will be responsible for creating a federated network that uses artificial intelligence to accelerate health care research in Spain. The project, known as TARTAGLIA, will take place under the R&D Missions in Artificial Intelligence program, which is part of the Digital Spain 2025 agenda and the National Artificial Intelligence Strategy. The project will be funded by the European Union using Next Generation EU funds, with a budget of over €7.5 million.
Clinical data associated with individuals is both critical and sensitive in nature, and for reasons of security and privacy, use of such data to create artificial intelligence models is highly restricted by law and by government policy. Application of new technologies to solve this problem, and the ability to experiment with governance of a secure federated network, are key factors that will allow all of Spain’s regions and government health institutions to contribute to the data banks, and also to benefit from them.
GMV’s work on the project will consist of applying advanced cryptographic methods to ensure that patient data remains encrypted while all of the necessary calculations are being performed, to create a balance between privacy and the ability to make use of that data, without exposing it and without the need to transfer it away from the organizations involved.
This work will enhance training of the mathematical models and support decision-making, while also contributing to the advancement of personalized, precision medicine. In turn, this will improve the treatment received by patients and accelerate the pace of clinical trials, among other benefits.
In addition to a line of research focused on a federated learning configuration, the TARTAGLIA project will use AI to address other research challenges, such as the use of ultrasound for diagnosis and applications for AI in four clinical areas: Alzheimer’s disease, prostate cancer, diabetes, and complex chronic disease.
As explained by Inmaculada Pérez Garro, director of Digital Health at GMV, “by creating a public-private consortium that brings together institutions and leading companies in their fields, TARTAGLIA will be a priority project for Spain, with a focus on driving innovation and knowledge throughout the country as well as internationally. We are very proud of this project because it received the highest score in this R&D program, and at GMV, research and innovation is part of our DNA.”
The consortium will be using the IBM Cloud Pak for Data solution to facilitate joint work by the data providers. This solution from IBM, which is supported by advanced analytical models and a data fabric, allows TARTAGLIA to apply AI on any of the Spanish and European Union systems where the data is housed, allowing the researchers to remain focused on their primary research objectives.
According to Juan Carlos Sánchez Rosado, IBM’s Health Industry Leader for Spain, Portugal, Greece, and Israel, “the project represents an excellent opportunity to make progress on developing federated artificial intelligence models, without any need to transfer data. This makes it easier to comply with the existing regulations and the need for data security, which is an especially critical element when sensitive information is involved. This in turn will allow for enhanced efficiency, flexibility, and automation when developing and applying AI models in the field of health care.”
Alfonso Valencia, director of the Life Sciences Department at the Barcelona Supercomputing Center agrees. “Demonstrating how AI approaches can function in the federated data environment represents a key stage for advanced use of biomedical information”, he says.
The TARTAGLIA project received financing from the Ministry of Economic Affairs and Digital Transformation (case no. MIA.2021.M02.0005), corresponding to funds from the Recovery, Resilience and Transformation Plan.
The public-private consortium includes 16 entities and has received the highest score of all projects submitted in the R&D Missions in Artificial Intelligence competition.
The project has a budget of over €7.5 million, funded through the European Union with Next Generation EU funds.