Last week researchers from IFCA-CSIC attended the AIHUB CSIC Summer School in Barcelona, Spain. Specifically, IFCA researchers presented three posters related to different research projects in the area of artificial intelligence. One of them, entitled “Application of federated learning to medical imaging scenarios” is part of the AI4EOSC project. One of the goals of AI4EOSC is to allow the creation of AI systems on distributed datasets, with a special focus on federated learning. The aim of the project is to support the management of experiments through the platform dashboard.
Some of the results obtained in the publication entitled “Study of the performance and scalability of federated learning for medical imaging with intermittent clients”, published in the journal Neurocomputing, and funded within the AI4EOSC project framework, were presented. The use case presented in the poster covers a medical image analysis of chest X-ray images using a federated learning architecture. Applying this technique to this use case is of particular interest since federated learning is a privacy-preserving data decentralization technique used to perform secure machine and deep learning. This allows collaboration between different data owners without sharing raw data, which in this area may be sensitive in nature. Stay tuned because soon we will publish a series of posts and tutorials about this topic!
This one-week conference (from July 3rd to 7th) brought together researchers from different fields and also teachers to share not only current research in the area of artificial intelligence and create synergies, but also to analyze together with educators ways of transferring this knowledge to the classroom.
The AIHUB CSIC network brings together more than 400 researchers from more than 80 research groups from 40 centers in Spain that apply AI in different research areas such as physics, microelectronics, life sciences, astronomy or philosophy.
