Medical AI
AI systems used in healthcare for diagnosis, treatment recommendations, patient monitoring, and clinical decision support.
Definition
Medical AI (also called Clinical Decision Support or Healthcare AI) refers to AI systems used in medical contexts—including diagnostic assistance, treatment recommendations, risk prediction, drug development, surgical robotics, and patient monitoring.
Regulatory Framework: Medical AI faces dual regulation under both AI-specific rules and medical device regulations:
- EU AI Act: High-risk under Annex III(5)(a) for AI "intended to be used as a safety component in the management and operation of... medical devices"
- EU MDR: Medical Device Regulation (2017/745) classifies AI as software medical devices
- US FDA: Regulates AI/ML-based Software as a Medical Device (SaMD) under 510(k), De Novo, or PMA pathways
Key Requirements:
- Clinical Validation: Evidence of safety and efficacy in clinical settings
- Transparency: Clinicians must understand AI recommendations' basis
- Human Oversight: AI supports rather than replaces clinical judgment
- Continuous Monitoring: Post-market surveillance for performance drift
- Data Quality: Representative training data across patient populations
FDA Categories:
- Locked algorithms: Don't change after deployment
- Adaptive algorithms: Learn and update over time (require special oversight)
Related concepts: Safety Component, Risk Management System, Post-Market Monitoring, Conformity Assessment
Sources
- •FDA Guidance on AI/ML
- •EU MDR
- •HIPAA
Related Terms
Safety Component
A part of a product or AI system that performs a safety function and whose failure could endanger people or property....
Risk Management System
Continuous process to identify, analyze, and mitigate AI system risks throughout its lifecycle....
Post-Market Monitoring
Ongoing surveillance of AI system performance after deployment in the market....
Conformity Assessment
Process to verify that high-risk AI systems meet regulatory requirements before market placement....