The Barcelona Supercomputing Center (BSC), the Faculty of Computer Science of Barcelona (FIB), and the Hackers@UPC organize Bitsxlamarató, a three-day hackathon to collaborate with the Marató of TV3, which this year is dedicated to COVID-19.
After a first edition dedicated to minority diseases, the organizers of the BitsxlaMarató want to bring together the efforts of teachers, research staff and any professional from the fields of health and technology (but also from other areas), who will work online as a team, during 3 days, to seek and develop solutions to face all the challenges posed by COVID-19.
The hackathon will take place on December 18-20 virtually and will include training talks. Participants who request it, will be able to use the MareNostrum supercomputer, both the general-purpose cluster and the MNCTE POWER9 cluster, ideal for tasks related to Artificial Intelligence.
On the first day, the organizers will pose to the participants a series of challenges in which they can be of help. During the next two days, the participants, organized into teams, will try to make their contributions to respond to these challenges. During the last day of the hackathon, the teams will make presentations and demonstrations of the solutions and solution proposals obtained to the challenges posed on the first day. There will be a prize for the best project of each challenge.
Challenge on the use of AI in the treatment of COVID-19 patients
The challenge in which the BSC is involved, together with the Language Technology Plan (TL Plan) of the Secretary of State for Digitization and Artificial Intelligence, is called "In search of similar patients: AI Doctor House conquers the Severe COVID-19 ". There is an urgent need on the part of healthcare professionals to more effectively access information that is relevant to clinical practice. From a COVID-19 point of view, Artificial Intelligence tools must distinguish between patients with and without risk of a serious outcome so that medical teams can intervene quickly. Specifically, this task aims to promote the development of systems capable of detecting similarities between a collection of texts of clinical cases.