Built for healthcare professionals

AI Fundamentals for Healthcare Professionals

Practical, safe, and responsible understanding of artificial intelligence in clinical care.

Start AI Foundations
12
Modules
20+
Knowledge checks
10min
Per topic
Abstract illustration of AI connecting healthcare data

Explore

My learning

Choose your learning pathway

Two pathways, each with its own certificate. Start with the foundations, then apply them in practice.

Beginner • 6 modules

AI Foundations

Build the essential knowledge needed to understand and responsibly use AI in healthcare.

Loading…
Intermediate • 6 modules

Applied AI in Healthcare

Apply AI literacy to model evaluation, clinical judgment, implementation, and ongoing monitoring.

Loading…
Specialty • Role-based

Specialty curricula

Standalone, role-based courses with their own progress tracking and certificate.

Complete Level 1 — AI Foundations first.

Why it matters

Why healthcare professionals need AI literacy

AI is already shaping how we document, diagnose, and decide. A working understanding keeps you — and your patients — safe.

Make informed decisions

Understand what AI tools can — and cannot — do before you trust them at the bedside.

Protect your patients

Recognize risks like bias, hallucinations, and privacy gaps before they reach care.

Lead your team

Speak the language of AI with confidence in rounds, committees, and vendor reviews.

Stay ahead of change

AI is being added to EHRs, imaging, and documentation. Don't get left behind.

Akron Children'sVisual learning library

Concepts you can see at a glance

Custom medical-illustration graphics — flat, clean, and made for clinical learners.

EXAMPLESMODELPREDICTIONSPattern matchedCurated, labeled dataLearns statistical patternsApplies to new cases

How AI learns from data

Examples in, patterns out, predictions applied to new cases.

Inpute.g. vitals, image pixelsHidden layersfind patterns·Outputprediction

Neural networks, simplified

Layered nodes that turn inputs into useful predictions.

ClinicianJudgment · Context · AccountabilityAI toolPattern recall · Speed · ScaleAI augments — the clinician decides.

Human + AI collaboration

AI augments; the clinician judges, decides, and signs.

TRAINING DATA1 pediatric exampleMODEL ACCURACYAdults93%Pediatric42%Underrepresented groups→ less reliable predictions→ ask before deploying

Bias in datasets

Models perform best on populations they were trained on.

CLINICIAN ASKS"Summarize thispatient's last visit."AI REPLIES✓ Reviewed for cough and fever.✓ Vitals stable, lungs clear.✓ Plan: supportive care, follow-up.⚠ Started on amoxicillin 500mg.(never actually prescribed — fabricated detail)← Hallucination: confident but falseAlways verify AI-generated clinical content against the source.

Hallucinations in generative AI

Confident-sounding output is not always correct — verify.

BEFORESee patientType notesSearch labsInbox triageLate chartingHours of admin · burnout riskWITH AI ASSISTSee patientAmbient scribeAuto-summaryInbox draft repliesReview & signClinician reviews every output · time back at the bedside

Clinical workflow augmentation

Less admin, more bedside — with review at every step.

SAFE AIPrivacyHIPAA-compliant toolsTransparencyShow what AI didValidationTested on your populationHuman oversightClinician signs offAccountabilityClear ownershipMonitoringWatch for drift & bias

Safe AI governance

Six pillars that keep clinical AI safe and accountable.

Try it yourself

Interactive AI simulations

Five short, hands-on demos. Educational only — no real patient data, no clinical decisions.

Good vs. poor prompts

Same model, two prompts. Specificity changes everything.

Poor prompt

"Write something about asthma for my patient."

Likely output: A generic, 800-word essay covering asthma history, pathophysiology, and adult treatment — not tailored, not printable.
Clear prompt

"Write a 5th-grade reading-level handout for a 10-year-old with mild persistent asthma. Cover daily controller, rescue inhaler, and when to call. Keep it under 200 words."

Likely output: A short, age-appropriate handout with three labeled sections and clear call-back triggers. Ready to edit and hand to the family.
Recipe: Tell the model who the reader is, what you want, the format, and the length.
How it works

AI assists clinicians — it doesn't replace them

The clinician stays at the center of every decision. AI is a tool that surfaces information, drafts notes, and flags patterns — humans interpret, decide, and care.

AI

Surfaces & summarizes

Pattern recognition, drafting, search, triage support.

Clinician

Interprets & decides

Judgment, context, communication, accountability.

Patient

Receives safer care

Better documentation, faster answers, human connection preserved.

Every recommendation from an AI tool is a starting point — never the final word. Human oversight is non-negotiable.

Try it now

Quick knowledge check

Four clinical scenarios. See how you'd respond — then learn why.

Knowledge check
Question 1 of 4

A patient asks if they can share their full medical history with a public chatbot. What's the best response?

Score: 0
Interactive game

AI myth vs. fact

Twenty statements about AI in healthcare. Decide whether each is a myth or a fact, then see how common the misconception is.

Question 1 of 20Score 0 / 20

"AI will replace physicians in most clinical roles within the next decade."

At the bedside

Clinical examples by role

Practical ways AI is showing up across hospitals, clinics, and pharmacies today. Filter by your role to see what's most relevant.

Ambient documentation

AI scribe drafts a SOAP note from your visit conversation so you can stay face-to-face with the patient.

PhysiciansAPPs
Safety note

Physician reviews, edits, and signs — never auto-publish.

Imaging triage assistance

AI flags a possible intracranial hemorrhage on a head CT and moves it to the top of the radiology worklist.

Physicians
Safety note

Decision support, not diagnosis. The radiologist still reads the study.

Evidence summarization at the point of care

Physician asks an institution-vetted tool to summarize current guidelines for a complex comorbid patient.

PhysiciansAPPs
Safety note

Cross-check citations. Generative tools can fabricate references.

Early deterioration alerts

Model flags an inpatient at rising risk of sepsis based on trending vitals and labs, prompting a bedside assessment.

NursesPhysiciansAPPs
Safety note

The alert is a prompt, not an order. Clinical judgment determines the response.

Medication reconciliation support

AI compares discharge meds against the active med list and surfaces discrepancies for the nurse to review.

NursesAPPsPhysicians
Safety note

Pharmacist or clinician confirms before any change is made.

Shift handoff summaries

AI drafts a concise SBAR-style handoff from the chart so the oncoming nurse starts with the key issues.

Nurses
Safety note

Outgoing nurse verifies accuracy before handing off.

Patient message drafting

AI suggests a draft response to a portal message in the patient's preferred language and reading level.

APPsPhysiciansNurses
Safety note

APP edits for accuracy and tone before sending.

Chronic disease follow-up prep

AI surfaces overdue screenings, abnormal trends, and missing labs before the visit so the APP can plan a focused encounter.

APPsPhysicians
Safety note

Use as a prep aid — confirm against the chart.

Learner concept explainers

Trainee uses an AI tool to explain an unfamiliar EKG pattern or ABG in plain language with worked examples.

Trainees
Safety note

Verify with a trusted reference and your supervising clinician.

Differential diagnosis brainstorming

Trainee enters a de-identified vignette to widen the differential before precepting with their attending.

Trainees
Safety note

Never paste PHI into non-institutional tools. AI suggestions are a starting point, not a plan.

Operational dashboards

AI surfaces throughput bottlenecks, no-show patterns, and staffing gaps across clinics to support weekly ops reviews.

Administrators
Safety note

Validate model assumptions and data sources before acting on recommendations.

Inbox and prior-auth triage

AI categorizes incoming faxes, referrals, and prior-auth requests so staff can route them faster.

AdministratorsNurses
Safety note

Spot-check categorization and monitor for drift over time.

AI governance and monitoring

Admin team tracks performance, bias, and drift metrics for each deployed clinical AI tool on a recurring cadence.

Administrators
Safety note

Pair quantitative metrics with frontline clinician feedback.

By the numbers

AI use in healthcare is accelerating

Whether or not you've chosen to use AI yet, it's increasingly part of your clinical environment.

79%

of healthcare organizations report using AI in some form

Source: HIMSS, 2024

60%+

of US clinicians have piloted ambient AI scribes

Source: Industry surveys, 2024

1,000+

AI/ML-enabled medical devices authorized by the FDA

Source: FDA, 2024

37%

projected annual growth of AI in healthcare through 2030

Source: Market analyses

Figures are illustrative ranges drawn from publicly available industry reporting and may vary by source and year.

My progress

Track modules, scores, and certificates

Loading your progress… Sign in to save your progress permanently across devices.

Help us improve the learning experience

Tell us what was unclear, what worked, or what you'd like to see next. Feedback goes straight to the education team.