Four new guides published for the development and regulation of AI in Health

The Health/AI Programme has published four new guides to facilitate the development, qualification and regulation of artificial intelligence tools in the healthcare field. These guides, prepared by the TIC Social Health Foundation, offer practical guidance for professionals and entities in the sector and include explanatory infographics to facilitate understanding.

The new guides address key aspects to ensure safe and effective use of AI in healthcare:

• Best Practices for the Development of Generative AI Tools in Healthcare.Large Language Models (LLMs)

Aimed at professionals and organisations that develop generative AI algorithms, this guide focuses on Large Language Models (LLMs) and provides recommendations to ensure their safe and ethical use. It includes fundamental principles and functionalities such as transparency, robustness and human supervision of the system.

• Application of European AI Regulations to Healthcare

It offers practical insight into how to interpret and apply the European AI Regulation (EU AI Act) in the healthcare field, detailing the risk classifications and obligations that must be fulfilled by all agents involved in the development and deployment of AI algorithms in the Catalan healthcare system.

• Qualification and classification of an IT application (based on AI) as a medical device in accordance with the European MDR/IVDR Regulations

It helps determine whether an AI application in Health is considered a healthcare product according to EU regulations and explains how to classify it correctly. It also analyses the implications of each risk class and the requirements necessary to comply with the regulation, referring to the European MDR and IVDR Regulations, as well as other complementary regulations.

• Guide to obtaining the CE marking of an IT application based on AI as a medical device in accordance with the European MDR/IVDR Regulations

It explains the steps necessary to obtain the CE marking, an essential requirement for the marketing and sale of AI applications in health within the European Union. It details the regulatory requirements that these systems must meet to guarantee their safety and effectiveness, following EU legislation and Spanish royal decrees on health products.

These guides are framed in the innovation axis of the Health/AI Programme, with the aim of supporting healthcare sector initiatives and ensuring that AI-based technologies are developed in a sustainable manner that is safe, reliable and aligned with current regulations.

For more information on support for AI initiatives, you can click on the following link.