• The directors of the Foundation and the Breast Area of the health group, Mercedes Gozalbo and Julia Camps, attend the meeting of the European project that develops a “swarm learning” to improve the accuracy of Artificial Intelligence.
  • The meeting, held in Athens, allows progress to be made on a common tool that will help to anticipate the disease, improve individualised treatments and monitor its evolution more precisely, reducing uncertainty for patients.   

Artificial Intelligence is not the future, it is the present. That is why the Ribera healthcare group has been part, for several months now, of the work team of the European ODELIA project, whose objective is to develop an advanced Artificial Intelligence model, including a secure exchange of data, in this case to create even more personalised predictive and follow-up models for breast cancer. This week, the project has taken a new step to advance a common tool, which will allow the exchange of valuable information without sharing patient data, with the working meeting attended by the directors of the Ribera Salud Foundation and the Breast Area, Mercedes Gozalbo and Julia Camps respectively. “The next step is for all members to start uploading images and data, to help the AI model learn,” explains the Foundation’s director.

The international prestige of the group’s Breast Area, led by Dr Camps, and the large number of patients treated in the breast units of Ribera hospitals, make Ribera a strategic partner of this European project. “ODELIA will help to improve the individualised treatment of patients and will also be a tool to help predict the disease. It will also reduce uncertainty about the evolution of breast cancer, thanks to the predictive model, as well as the response to drugs, which will allow patients’ treatment to be personalised earlier and with greater certainty,” explains Dr. Julia Camps.

The ODELIA project is a decentralised swarm machine learning solution that allows multiple collaborators to share information without sharing the patient data itself, protecting the privacy and security of the data, yet allowing all collaborators to benefit from the collective learning. Data governance and privacy are preserved.

“When organisations only have access to their own data, their AI models evolve based only on information about the people the organisation has worked with or works with, which creates a bias in the models. With swarm learning, an organisation can combine its data with learnings from other organisations, increasing accuracy and reducing bias,” explains the director of the Ribera Salud Foundation.

The next ODELIA project meeting will take place in spring.