#5 Newsletter: Standardization

Several activities on Standardization and Legislation have been carried out in Task 6.3, led by EFMI.

In the field of standardization, monitoring the activities in various standards development working groups and analyzing their relationship with the project has been a key activity. In particular, two major ISO committees are taking important steps in the standardization of AI in Health: ISO/TC 215 (Health Informatics) and ISO/IEC JTC1/SC42 (Artificial Intelligence).

Concerning ISO/TC 215, after several initial steps, a Task Force ISO/TC 215/TF 5 on “AI technologies in health informatics” has been created. On the other hand, a Joint Working Group ISO/IEC JTC1/SC42 – ISO/TC 215 on “AI-enabled health informatics” (JWG3) has also been created and just started to work. EFMI members are participating in both groups.

As current work in Task 6.3, the new documents being developed by these standardization groups, such as “Use Cases” from JWG3  and “Application of AI technologies in health informatics” in TF5, are being mapped with the results of the HosmartAI project to identify potential contributions to standardization.

In the field of legislation, national, regional, and local legislation applicable to the pilots has been identified and analyzed, generating a report with recommendations of interest to all pilots to reinforce the necessary fulfillment of the European legal framework.

On the other hand, work is being done on applying the FAIR principles to the pilot data collections and assessing the maturity level of such fulfillment through adapting the FAIR maturity model of the Research Data Alliance (RDA). Two workshops have been developed to raise awareness on this issue for the pilot teams, one online and one in person, during the Eindhoven plenary on November 22, 2022.

Work is currently underway on developing the maturity model adapted for pilot data and on a scientific paper on relevant aspects for research on applying FAIR principles to data to improve the performance of AI use in Health, based on the HosmartAI experience.