Can we trust AI in medical devices?
20 November 2023
Artificial intelligence (AI) and machine learning (ML) has the potential to revolutionise medical devices, and develop solutions that can scale and increase efficiency in medical care.1 However, how can we keep up with and control the rapid post market learning of these algorithms to ensure that the quality and safety of these devices are not compromised and can be regularly monitored?
Currently, there are no laws or harmonised standards that specifically regulate the use of artificial intelligence in medical devices. There has been much research, discussion, and guidance regarding how to manage regulatory compliance by the Medicines and Healthcare products Regulatory Agency (MHRA) in the UK1, focused on challenges in the regulatory landscape including post marketing compliance and how to manage this at scale.
The trouble is, with continuous machine learning, can the device change its behaviour to a point where it needs resubmission for approval? How do you quantify changes in what maybe effectively become a black box, where the logical processes and interactions are not immediately obvious to an outside human observer?
To address this risk, the MHRA recommends that regulatory guidelines for the development 1, validation and use of artificial intelligence with medicinal products should be developed in areas such as data provenance, reliability, transparency and understandability, pharmacovigilance, and real-world monitoring of patient functioning. These guiding principles will help promote safe, effective, and high-quality medical devices that use artificial intelligence and machine learning.
There is a fine balance where too much regulation would limit the benefits of innovation, whereas insufficient regulation could allow to inadequate or unsafe devices onto the market2. Therefore, it is crucial to strike a balance between innovation and safety.
As medical device developers, we are very used to working within the relevant regulatory frameworks. We support the discussions between regulatory bodies and industry on how to adequately regulate the use of artificial intelligence in medical devices.
When they are created, new guidelines will help promote safe, effective, and high-quality medical devices that use artificial intelligence for the benefit of patients and wider society. We’ll be following the updates closely to ensure we can apply any regulations to our projects accordingly.
Contact us for help with your medical device design.
Written by Alex Vasiev, Springboard Engineer.
1: Software and Artificial Intelligence (AI) as a Medical Device – GOV.UK (www.gov.uk)
2: The Regulation of Artificial Intelligence as a Medical Device (publishing.service.gov.uk)
