Share your views: Draft Pre-market guidance for machine learning-enabled medical devices

From Health Canada

With the increase in artificial intelligence and machine learning in medical devices, Health Canada has developed the Draft Pre-market guidance for machine learning-enabled medical devices. This guidance:

• helps manufacturers who submit an application for a machine learning-enabled medical device (MLMD)
• outlines our expectations for demonstrating the safety and effectiveness of an MLMD throughout its lifecycle
• introduces a mechanism to pre-authorize planned changes to an MLMD to address risks through a pre-determined change control plan (PCCP)

Join in: how to participate

Send an email to mddpolicy-politiquesdim@hc-sc.gc.ca with your ideas or comments to make yourself heard.

Who is the focus of this consultation

We seek to engage with medical device stakeholders, including:

• manufacturers of Class II to IV medical devices
• regulatory representatives
• machine learning experts

Key questions for discussion

Health Canada invites you to provide feedback on the overall clarity and content of the draft guidance.

We also would like you to answer the following questions:

• Do we need to add or remove information on the device's intended use or description?
• Do you agree that risk analyses are important in determining if all applicable risks are mitigated appropriately? Do you support their evalution by Health Canada? Should other risks be listed?
• Data selection, training and testing must be done correctly to ensure appropriate performance in the real world. Are the data selection and training elements appropriate? Should other considerations be made?
• Are the minimum clinical evidence requirements reasonable for this new technology? Are there challenges in demonstrating the impacts on different sexes, genders and under-represented populations?
• Transparency of any machine learning system is important to Health Canada. Users should know what the system is meant to do, the composition of its training data, and its performance and limitations. What information should we add or remove?
• What information should we add or remove from the PCCP sections?

Health Canada will use the comments we receive to finalize the guidance document. Your feedback will help us clarify the factors that applicants should consider to demonstrate that the device is safe, effective and of high quality throughout its lifecycle.

Related information

• Food and Drugs Act
• Medical Device Regulations
• Guidance on supporting evidence to be provided for new and amended licence applications for Class III and Class IV medical devices, not including in vitro diagnostic devices (IVDDs)
• Draft Health Canada IMDRF table of contents for medical device applications
• IMDRF N67: Machine learning-enabled medical devices: Key terms and definitions
• Software as a medical device (SaMD): Definition and classification
• Good machine learning practices for medical device development: Guiding principles

Contact us

Medical Devices Directorate
Health Products and Food Branch
Health Canada
11 Holland Ave, Tower A
Address Locator: 3002A
Ottawa ON K1A 0K9

Email: mddpolicy-politiquesdim@hc-sc.gc.ca