AI & Clinical Decision Support

AI and machine learning in clinical practice. Covers clinical decision support systems, FDA regulation of AI/ML medical devices, algorithmic bias, liability and standard of care with AI, and the future of AI-augmented medicine.

6 Units

20 minutes per unit

Curriculum Map

What You Will Learn

FDA AI/ML Framework

SaMD classification, 510(k) vs PMA pathways, predetermined change control plans, and post-market surveillance requirements.

Bias & Equity

Training data bias, racial and gender disparities in algorithms, fairness metrics, and evidence-based mitigation strategies.

Liability

Physician responsibility when relying on AI output, the black box problem, informed consent obligations, and malpractice implications.

All Units

20 minutes

AI and Machine Learning in Healthcare: Foundations and Applications

Examines the technical foundations of AI and machine learning in healthcare, including learning paradigms, neural network architectures, and current clinical applications.

•Distinguish between supervised and unsupervised learning methodologies in clinical contexts
•Identify how neural networks process medical imaging and diagnostic data
•Evaluate current use cases for machine learning algorithms in patient care delivery

Start learning

20 minutes

Clinical Decision Support Systems: Design, Deployment, and Integration

Examines clinical decision support system architectures, evidence-based alert frameworks, medication safety interventions, and workflow integration strategies.

•Categorize types of clinical decision support and their intended use cases in care delivery
•Evaluate evidence-based alert systems for medication safety and diagnostic assistance
•Analyze workflow integration challenges and strategies for effective CDS implementation

Start learning

20 minutes

FDA Regulation of AI/ML Medical Devices: Pathways and Oversight

Examines FDA regulatory frameworks for AI/ML medical devices, including SaMD classification, 510(k) and PMA pathways, and the predetermined change control plan approach to algorithm updates.

•Distinguish between 510(k) clearance and PMA approval pathways for AI/ML-based medical devices
•Identify how software as a medical device (SaMD) classification affects regulatory requirements
•Evaluate the FDA's predetermined change control plan framework for continuously learning algorithms

Start learning

20 minutes

Algorithmic Bias and Health Equity: Sources, Detection, and Mitigation

Examines sources of algorithmic bias in healthcare AI, documented disparities in algorithm performance, fairness measurement frameworks, and strategies for mitigating bias in clinical deployment.

•Identify how training data composition introduces racial and gender disparities in algorithm performance
•Evaluate fairness metrics and their limitations in measuring algorithmic equity
•Apply evidence-based mitigation strategies to reduce bias in clinical algorithm deployment

Start learning

20 minutes

Liability and Standard of Care with AI: Physician Responsibility in the Age of Algorithms

Examines medical malpractice implications of AI use in clinical practice, allocation of liability between clinicians and device manufacturers, the black box interpretability problem, and evolving standards of care.

•Analyze how physician liability is allocated when diagnostic or treatment decisions involve AI tools
•Evaluate the legal implications of relying on or deviating from AI-generated recommendations
•Identify how the black box problem affects the standard of care and informed consent obligations

Start learning

20 minutes

The Future of AI in Clinical Practice: Transformation, Competency, and Accountability

Examines emerging applications of AI in autonomous diagnosis, precision medicine, and workflow transformation, and identifies professional competencies required for responsible AI integration in clinical practice.

•Evaluate the feasibility and implications of autonomous diagnostic systems in clinical deployment
•Analyze how AI-enabled personalized medicine will transform treatment selection and monitoring
•Identify core competencies clinicians will need to work effectively with AI tools

Start learning