Standards Development

Scope

This report provides guidance for applying an ISO 14971 risk management process when evaluating medical technology utilizing artificial intelligence and, in particular, machine-learning. The report addresses how hazards already identified in ISO 14971 could affect the safety and effectiveness of medical technology incorporating machine-learning AI, provides examples, and suggests strategies for eliminating or mitigating the associated risk. The report also explores additional unique or emergent hazards/hazardous situations. This report does not modify the risk management process—rather it provides information and guidance to inform the application of ISO 14971 to AI/ML medical technology.

Purpose

The purpose of this standard is to explore potential new hazards for AI and provide examples of those hazards, so that these can be incorporated into manufacturer’s systems for assessing and managing risk for Medical Device Learning Systems. The standard will provide initial consensus guidance to assist those who are applying ISO 14971 to regulated AI medical technologies.
ISO 14971 provides a methodology for assessing and managing the risk associated with medical devices. It has been recognized by medical device regulators and adopted as a national standard in countries across the world.  It is the global standard used by medical device manufacturers and regulators to govern risk management  
BSI and the US Standards Development Organisation, the Association for the Advancement of Medical Instrumentation (AAMI) are working together to develop guidance on the use of AI as a medical technology. A series of joint workshops with stakeholders in the US and UK have been held, along with publishing two whitepapers on the topic. One of the key outcomes from this work was the recommendation to develop “risk management guidance to assist in applying ISO 14971 to AI as a medical technology”.
This standard will be a jointly developed and published in partnership with AAMI. The ambition for this guidance is to become the basis ISO/IEC guidance on risk management of AI as a medical technology in the future and may be the first of a series of jointly developed work items by AAMI and BSI. The BSI drafting panel will work in conjunction with the AAMI drafting panel to agree a common text that will be published by AAMI and BSI. For simplicity purposes, BSI will publish this joint work as a British Standard in the UK and AAMI will deliver a Technical Information Report (TIR) for the US market. A memorandum of understanding will be developed between BSI and AAMI to support the work.

There is currently no guidance that explains the particular hazards for AI as a medical technology. ISO 14971 is a widely-used risk management standard that is aligned to global regulatory frameworks. This guidance will identify the particular hazards relating to AI that need to be considered in ISO 14971.