An analysis grid for evaluating AI-embedded medical devices
Regulations
13/04/2022
The HAS has updated its filing guides to take into account the development of artificial intelligence (AI) in the medical devices it evaluates. An analysis grid is available to guide manufacturers.
Promoting patient access to innovation
The descriptive grid of functionalities of medical devices embedding a system with machine learning is a framework for the evaluation of medical devices embedding artificial intelligence. It enables manufacturers to provide information adapted to all medical devices incorporating machine learning, whatever the technique used: data used for the algorithm, types of learning, performance thresholds, etc.
.
By following this mapping (42 items, 4 key areas covering the entire development process of a technology), they can objectify the characteristics of their medical device, and thus provide the HAS with all the elements necessary for the assessment of their product.
This tool has been directly integrated into the guide de dépôt de dossiers with a view to device reimbursement.
Also available in English: grille d'analyse / guide de dépôt.
Find out more
Artificial intelligence and healthcare technologies: the challenge of evaluation
How should we think about the assessment of healthcare technologies using artificial intelligence? How can we exploit these data qualified as high-risk? What are the prospects at European level? Elements of an answer with Isabelle Adenot (president of the national commission for the evaluation of medical devices and health technologies - CNEDiMTS, member of the HAS College) and Bertrand Braunschweig (scientific coordinator of the Confiance.ai program), on the occasion of the HAS-PFUE colloquium on March 7, 2022.
