Transforming healthcare: balancing innovation and compliance in AI-driven medical devices

The healthcare landscape is in the process of being dramatically transformed by the advent of digital health technology, driving innovation and revolutionizing healthcare delivery. Artificial Intelligence (AI) and Machine Learning (ML) are particularly poised to impact several healthcare domains, from diagnostics and treatment to patient monitoring and management. These technologies hold significant promise, especially in underfunded medical research areas, by aiding in the detection of comorbidities and identifying environmental or genetic factors that increase disease risk.

AI and ML in healthcare delivery

While concerns about potential harm and bias are carefully considered, the development of AI/ML tools in healthcare is accelerating. These tools are designed to manage and interpret vast amounts of data quickly and accurately, whether in text, video, or imagery. By automating data analysis, AI/ML can save countless hours of manual review and provide insights that would take human reviewers years to uncover.

As these advancements take hold, regulatory bodies are grappling with the complexities of evaluating and approving non-traditional medical devices, particularly those without a physical presence. In the US, UK, and EU, regulatory progress is being made to ensure the safety and efficacy of AI/ML-enabled medical devices. The evolving regulatory landscape, including new regulations like the EU AI Act, presents challenges, especially as digital health solutions increasingly blur the lines between medical and non-medical tools.

In fact, differentiating between digital health and digital medical devices is not always straightforward. Digital health technologies range from non-medical devices designed for well-being to specific medical devices. The International Medical Device Regulators Forum (IMDRF) provides a consensus definition for “Software as a Medical Device” (SaMD) as "software intended to be used for one or more medical purposes that perform these purposes without being part of a hardware medical device."

Regulatory approaches in the UK, US and EU

In June 2023, the UK's MHRA confirmed it is developing guidance to help clearly identify SaMD, distinguishing them from other tools such as wellbeing software, lifestyle products, IVD software, and companion diagnostics[1].

In the same year, the MHRA also updated its Software and AI as a Medical Device Change Programme to clarify future regulatory requirements and protect patients. This program aims to position the UK as a leader in responsible innovation for medical device software, establishing safety assurances, clear guidance, and collaborative efforts with key partners like NICE and NHS England, as well as international regulators through the IMDRF.

Additionally in July 2024, the UK will launch its new regulatory sandbox pilot, providing a safe space for AI tool developers in healthcare to trial innovations under regulatory scrutiny[2].

In Europe, the final text for the AI Act, published this year, will regulate AI systems across industries, including healthcare. This act underscores the importance of reliability and accuracy in AI diagnostics and decision-support systems, especially when classifying high-risk AI systems.

As a result of wider definitions of high-risk AI uses, certain digital health products may also come under CE marking regulation for the first time under AI Act regulatory assessment of high-risk AI systems, where they are not considered to be medical devices. These digital health products will therefore now require notified body assessment and CE marking as an AI system, which will be based on different criteria to existing requirements for medical devices.

In the US, the FDA has extensive experience regulating AI/ML-enabled devices and maintains a publicly available list of AI tools with FDA marketing clearance[3]. The FDA also accounts for the need of adaptive re-training of certain AI/ML systems, allowing pre-authorized software changes without further regulatory assessment. This innovative approach contrasts with traditional EU methods, which can struggle with similarly effective regulation.

The digital health landscape continues to evolve, and navigating the complexities of AI-based medical devices requires collaborating with regulatory experts and implementing a multidisciplinary approach. This will ensure compliance with evolving standards, facilitate market entry, and promote the successful integration of innovative technologies in healthcare.

[1] Gov.co.uk, Software and AI as a Medical Device Change Programme – Roadmap, June 2023

[2] Gov.co.uk, MHRA launches AI Airlock to address challenges for regulating medical devices that use Artificial Intelligence, 9^th May 2024

[3] FDA, Artificial Intelligence and Machine Learning (AI/ML)-Enabled Medical Devices