On 10 June 2024, the COMFORT project participated in a successful roundtable discussion alongside other leading EU-funded projects from topic HORIZON-HLTH-2022-TOOL-12-01-two-stage "Computational models for new patient stratification strategies". The event, hosted by Aristotle University of Thessaloniki (AUTH) and organised by the PREPARE project, brought together representatives from Microb-AI-ome, SPIDeRR, STRATA-FIT, STRATIFAI, TARGET, AI4Lungs, CoMPaSS-NMD, and COMFORT.

The roundtable looked at common challenges faced by AI-focussed healthcare projects. Discussions addressed crucial topics like data interoperability, federated learning approaches, Explainable AI tools, and navigating Medical Device Regulation (MDR) certification.

COMFORT was well-represented by both Antonis Billis (AUTH), who participated in person, and project coordinator Keno Bressem, who joined remotely. Keno Bressem provided a detailed introduction to COMFORT, highlighting its scope, objectives, and the use of multimodal data to develop AI models. He also emphasised the importance of common data formats for successful implementation. Antonis Billis further contributed by outlining COMFORT's plans for achieving certification of its AI models.

The roundtable discussions highlighted key considerations for all AI healthcare projects. These include being mindful of data bias sources, maintaining detailed technical documentation, implementing version control for AI models, and establishing robust quality assurance processes, particularly for labelling procedures in supervised learning models.

COMFORT's participation in this collaborative event signifies the project's commitment to fostering knowledge exchange and best practices within the field of AI for healthcare. The insights gained will undoubtedly contribute to COMFORT's as well as all other participating projects’ ongoing development and success.