Summary
Health Canada guidance for manufacturers submitting new or amendment applications for Class II, III, and IV machine-learning-enabled medical devices. The guidance explains MLMD scope, risk classification, lifecycle evidence expectations, good machine learning practice, design, risk management, data selection and management, development and training, testing and evaluation, clinical validation, transparency, post-market monitoring, and predetermined change control plans.
Healthcare Implications
Canadian sponsors and manufacturers of ML-enabled medical devices should structure premarket submissions and lifecycle documentation around Health Canada’s expectations for safety and effectiveness, representative datasets, SGBA Plus, transparency, PCCPs, change management, and post-market monitoring. Health systems and purchasers can use the guidance as a reference for vendor diligence and deployment-risk review.