Newton’s Tree has been selected by the Medicines and Healthcare products Regulatory Agency (MHRA) as a partner of the AI Airlock programme. AI Airlock, a first of its kind regulatory sandbox, was established in partnership with the Department of Health and Social Care, NHS AI Lab, and Team AB, to address regulatory challenges faced by AI as a Medical Device.
As part of AI Airlock, Newton’s Tree will further develop its Federated AI Monitoring Service (FAMOS), a key component of the Newton’s Tree platform. FAMOS gives hospitals, AI developers, and regulators the ability to observe and monitor AI performance in near real time, which helps them to tackle the challenge of drift; where changes in clinical environments, input data or human-computer interactions can degrade AI performance. This is a growing concern as the UK accelerates its uptake of healthcare AI.
Appropriate monitoring of medical AI for drift currently has relatively little formal guidance globally, especially considering the probability and impact of occurrence. This is a problem because it means it is unclear what best practice looks like for manufacturers and users. AI Airlock is an opportunity to address that problem, which is a necessary step if we are going to turn the national ambition of safe medical AI into a reality.
- Robin Carpenter, Head of AI and Governance and Policy at Newton’s Tree
The AI Airlock will have several key deliverables across the next six months of the programme:
● Monitoring drift: validating FAMOS as an approach to monitoring drift.
● Strengthening regulatory pathways: collaborating with the MHRA, AI vendors, and other key partners to explore how monitoring can support effective risk management frameworks.
● Defining key responsibilities: clarifying the roles, responsibilities and expectations of key stakeholders when monitoring AI.
FAMOS is supported with funding from InnovateUK and is being conducted in collaboration with Leeds Teaching Hospitals NHS Trust, NHS Greater Glasgow and Clyde, Annalise-ai, Infervision, Qure.ai, Imperial College London, and Vita Healthcare Solutions, with support from NVIDIA.
Effectively managing risks associated with an AI medical device throughout its lifecycle is crucial for everyone involved, from the developers, to healthcare professionals, to the patient. I’m excited about this opportunity to improve best practices and advance healthcare.
- Mark Philips, Chief Clinical Officer, Annalise.ai
Partnering with Newton’s Tree and the MHRA AI Airlock will allow us to explore how continuous monitoring can elevate the safety of AI across the NHS. This is a crucial step in ensuring these technologies are used responsibly and ultimately benefit patients.
- Professor David Lowe, Director of Clinical Innovation at University of Glasgow and Emergency Consultant at NHS Greater Glasgow and Clyde.
Ensuring AI applications remain safe and effective after they are deployed is critical for both hospitals and vendors. Get in touch if we can support your team with keeping AI applications safe within your organisation.