AI in Clinical Research

An introduction to artificial intelligence concepts, written for clinicians rather than engineers. This section answers the question: what is AI, and what can it do for my research?

• 1.1What Is Artificial Intelligence? A Clinician's Primer
• 1.2Supervised, Unsupervised, and Self-Supervised Learning
• 1.3Common AI Architectures in Medicine
• 1.4Foundation Models and Large Language Models in Clinical Research
• 1.5Matching the Problem to the Approach
• 1.6What AI Cannot Do: Limitations and Common Misconceptions

Building effective AI models begins with data. This section covers everything a clinical researcher needs to know about assembling, preparing, and safeguarding datasets.

• 2.1Why Data Quality Matters More Than Model Complexity
• 2.2How Much Data Do You Need?
• 2.3Data Annotation and Labeling for Clinical AI
• 2.4Class Imbalance: When Rare Conditions Dominate Your Research Question
• 2.5Data Preprocessing and Augmentation
• 2.6Multi-Site and Multi-Device Data: The Generalization Challenge
• 2.7De-Identification, Privacy, and Regulatory Compliance for Research Data
• 2.8Synthetic Data and Data Simulation

A practical guide to the model development lifecycle, written for clinical researchers who will collaborate with engineering teams rather than build models alone.

• 3.1The AI Development Lifecycle: From Hypothesis to Deployment
• 3.2Training, Validation, and Test Sets: Why You Need All Three
• 3.3Transfer Learning: Standing on the Shoulders of Large Models
• 3.4Hyperparameter Tuning and Model Selection
• 3.5Overfitting, Underfitting, and Regularization
• 3.6Working with an Engineering Team: The Clinician's Role
• 3.7Tools and Infrastructure: What You Need to Know

Technical metrics for assessing how well an AI model performs its intended task. This section gives clinical researchers the vocabulary to critically evaluate published results and their own models.

• 4.1Classification Metrics: Sensitivity, Specificity, and Beyond
• 4.2The ROC Curve and AUC: What They Tell You and What They Hide
• 4.3Segmentation Metrics: Dice, IoU, and Volumetric Measures
• 4.4Calibration: Does a 90% Confidence Score Really Mean 90%?
• 4.5Statistical Significance, Confidence Intervals, and Sample Size for AI Studies
• 4.6Cross-Validation Strategies for Medical Data
• 4.7Comparing Models: When Is One Model Truly Better Than Another?

Moving beyond technical performance to demonstrate that an AI model provides real clinical value. This section bridges the gap between model development and clinical evidence.

• 5.1Technical Validation vs. Clinical Validation: Understanding the Difference
• 5.2Designing Retrospective Validation Studies
• 5.3Designing Prospective Clinical Validation Studies
• 5.4Multi-Site External Validation
• 5.5Reader Studies: Measuring AI's Impact on Clinical Decision-Making
• 5.6Defining Clinically Meaningful Endpoints
• 5.7Reference Standards and Ground Truth in Medical AI

Navigating the regulatory landscape for bringing an AI-based medical device to market. This section covers U.S. FDA requirements in depth, with orientation to international frameworks.

• 6.1When Does Your AI Algorithm Become a Medical Device?
• 6.2Device Classification: Class I, II, and III
• 6.3The 510(k) Pathway: Substantial Equivalence to a Predicate
• 6.4The De Novo Pathway: First-in-Class Devices
• 6.5Premarket Approval (PMA) and When It Applies
• 6.6Breakthrough Device Designation: Faster Review for Innovative AI
• 6.7Good Machine Learning Practice (GMLP): The 10 Guiding Principles
• 6.8Predetermined Change Control Plans (PCCP)
• 6.9Software Lifecycle Requirements: IEC 62304
• 6.10Design History File and Quality Management System Essentials
• 6.11International Regulatory Frameworks: EU MDR, Health Canada, and Beyond
• 6.12Post-Market Surveillance and Real-World Monitoring

Ensuring that AI systems are equitable, transparent, and trustworthy. This section addresses the scientific, ethical, and increasingly regulatory dimensions of responsible AI development.

• 7.1Sources of Bias in Medical AI
• 7.2Demographic Fairness: Performance Across Populations
• 7.3Generalization Across Devices, Sites, and Geographies
• 7.4Explainability and Interpretability in Clinical AI
• 7.5Transparency and Reproducibility
• 7.6Ethical Frameworks for AI in Healthcare

Getting an AI tool from the lab into clinical practice. This section covers the practical, technical, and human factors that determine whether a validated model actually improves care.

• 8.1From Model to Product: The Implementation Gap
• 8.2Clinical Workflow Analysis: Where Does AI Fit?
• 8.3Interoperability: DICOM, HL7 FHIR, and Integration Standards
• 8.4Human-AI Interaction and User Interface Design
• 8.5Clinician Trust, Adoption, and Change Management
• 8.6Monitoring AI Performance in Production

Standards and best practices for communicating AI research to the clinical community. This section helps researchers produce publications that are rigorous, reproducible, and useful.

• 9.1Reporting Standards: CONSORT-AI, SPIRIT-AI, TRIPOD+AI, and STARD
• 9.2Writing the Methods Section for an AI Study
• 9.3Common Methodological Pitfalls in Medical AI Papers
• 9.4Sharing Code, Data, and Models
• 9.5Peer Review of AI Manuscripts: What Reviewers Look For

• AGlossary of AI and Machine Learning Terms for Clinicians
• BRecommended Reading and Courses
• CRegulatory Standards Quick Reference
• DChecklist: Is Your AI Study Ready for Publication?