Skip to content

stanfordmed/aimeded

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

38 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Explore AI Diagnostic Performance

An interactive educational tool for teaching clinicians how to interpret AI diagnostic metrics using a chest X-ray pneumonia detection example.

Overview

This tutorial helps clinicians understand:

  • Confusion Matrix: True positives, false positives, true negatives, false negatives
  • Key Metrics: Sensitivity (recall), specificity, precision (PPV), accuracy, F1 score
  • ROC and PR Curves: How threshold selection affects the tradeoff between metrics
  • Prevalence Effects: Why the same AI performs differently in high vs. low prevalence settings

Features

  • Interactive threshold slider to explore metric tradeoffs
  • Three clinical scenarios: Severe Cases, Subtle Cases, and Low Prevalence
  • Toggle between ROC and Precision-Recall curves
  • Adjustable prevalence slider to demonstrate Bayesian effects on precision
  • Clinical scenario questions with expandable answers

Usage

Open index.html in a web browser. No build step or server required.

Authors

  • Vishnu Ravi, MD
  • Alaa Youssef, PhD
  • Aydin Zahedivash, MD
  • Gabriel Tse, MD
  • Jonathan Chen, MD, PhD

License

Stanford Medicine - AI in Medical Education

Links

About

Stanford AI in Medical Education Open-Access Tools

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published