Since 2018, Health Canada has undertaken an initiative to adapt its regulatory approach to better support digital health technologies, in particular medical devices. The main focus areas include artificial intelligence, software as a medical device, cybersecurity, medical device interoperability, wireless medical devices, mobile medical applications and telemedicine. To achieve this goal, Health Canada created the Digital Health Division under the aegis of the Bureau of Medical Devices and intensified its efforts to develop internal expertise.

On October 27, 2021, Health Canada, the United States Food and Drug Administration (FDA) and the United Kingdom Medicines and Healthcare Products Regulatory Agency (MHRA) jointly released the document Good Machine Learning Practice for Medical Device Development: Guiding Principles. The document consists of 10 guiding principles to help promote the safe, efficient and high-quality use of artificial intelligence and machine learning (AI / ML) in medical devices. Artificial intelligence is a field that combines computing and data to enable machines to mimic human intelligence when solving problems. Machine learning is a subfield of artificial intelligence where algorithms are designed to learn and improve a machine through experience (i.e. new data). These algorithms can be “locked”, so that their function does not change, or “adaptive”, which means that their behavior can change over time.

AI / ML medical devices have the potential to revolutionize the healthcare industry as they continuously learn and improve their performance using real world data after their deployment in the market. However, such medical devices are difficult to regulate, as Health Canada needs a mechanism to ensure safety and efficacy as medical devices “adapt” over time.

The current regulatory approach used by Health Canada to approve medical devices is not suited to the iterative and data-driven nature of AI development. Health Canada faces various challenges in developing an appropriate framework for AI / ML medical devices, including:

• fostering innovation without sacrificing safety and efficiency;
• establish the manufacturer’s requirements for pre-market authorization;
• ensure that the data sets used during development are reliable and representative;
• define ideal performance measures to correctly assess the performance of the algorithm;
• ensure that the integration of AI is done appropriately without unintended consequences;
• developing a framework for post-market regulation; and
• clarify the rules of liability if errors are made by the software.

The guiding principles mark Health Canada’s ongoing effort to adjust its approach to better regulate and facilitate market access for adaptive AI / ML medical devices.

The ten guiding principles are:

1. Leverage multidisciplinary expertise throughout the product lifecycle to ensure safe and effective AI / ML medical devices.
2. Implement good software and security engineering practices, including good data quality assurance, data management, cybersecurity, and orderly risk management and design processes.
3. Using clinical study participants and data sets that are representative of the target patient population (factors include age, sex, gender, race and ethnicity).
4. Use of training data sets independent of test data sets. Training data sets refer to data used to train a medical device, while test data sets refer to data used to test a medical device after it has been trained to measure performance, accuracy and efficiency.
5. Develop reference datasets based on the best available methods to ensure that clinically relevant and well characterized data are collected and that the limitations of the reference datasets are understood.
6. Adapt the model design to the available data and ensure that the model design reflects the intended use of the device.
7. If humans are involved in the model, pay special attention to human AI performance rather than isolated AI performance.
8. Develop and use robust test plans that demonstrate device performance under clinically relevant conditions.
9. Make available and deliver clear and contextually relevant information to users, including healthcare providers and patients.
10. Monitor performance and manage re-conversion risks after model deployment.

Although there are active AI / ML medical devices approved by Health Canada on the market today, these devices use locked-in rather than adaptive AI algorithms. The table below illustrates some of the AI ​​/ ML medical devices that have been approved by Health Canada.

These guiding principles are the latest example of Health Canada taking steps to modernize its regulatory approach to emerging technologies such as AI / ML medical devices. These steps are essential to foster AI innovation in healthcare without sacrificing the safety and efficiency of medical devices for its users. While challenges remain, it is encouraging to see Health Canada taking a systematic approach and working with its international counterparts.

[View source.]