TARTAGLIA – Federated Network of Artificial Intelligence to accelerate Health Research

AI has the potential to improve people's health, speed up clinical and biomedical research to answer scientific questions and challenges in the healthcare sector, provide more efficient healthcare services and substantially reduce costs. However, the development of robust and accurate AI models requires access to large databases with an appropriate volume of clinical information. The Spanish and European institutions have large data repositories, but they are isolated and not linked to each other.

The TARTAGLIA project was created with the aim of breaking down the data silos of healthcare centers through the creation of a federated learning network that allows sharing this information while preserving its privacy and confidentiality in order to obtain, in a common way, an AI model.

Tartaglia's objectives

• Accelerate the application of AI in the Spanish Healthcare System.
• Help patients and health professionals to reinforce prevention and early diagnosis of diseases.
• Create an ecosystem that channels research activity, makes it possible to build AI tools based on data shared securely, and create a competitive market in Spain and Europe to encourage the creation and growth of companies.

A federated data space

The creation of a network with a federated data space will make it possible to share health data without having to leave the repositories of the health centers that keep them, making the most of the potential of data from different regions and the interconnection with European projects.

In this way, an intelligent data analysis system will be created that allows anticipating and acting early, supporting decision-making, contributing to personalized and precision medicine, improving patient treatment and accelerating clinical trials, among others. For this, the Tartaglia project, beyond creating this federated learning network, focuses on the application of Artificial Intelligence in four clinical areas: Alzheimer's, prostate cancer, diabetes and complex chronic diseases.

The Tartaglia project at Ace

As part of the Tartaglia project, at Ace we study spontaneous speech through AI in order to identify parameters that allow us to predict the onset of Alzheimer's disease.

Our experience in the field of Alzheimer's and other dementias has led us to observe that language is a fundamental variable when it comes to identifying the first symptoms of the disease. However, in order to analyze a natural conversation, we need very complex algorithms and computational resources to process the language and study different parameters: syntax, morphology, lexicon, sound, etc.

Trataglia opens the doors to the world of Artificial Intelligence through the Accexible digital platform, which allows us to automatically process conversations we have with our patients. Thanks to this, we can exhaustively evaluate spontaneous speech in search of signs of Alzheimer's onset in order to make an early diagnosis.