Welcome to the collaborative curriculum repository for the Healthcare AI Micro-Credentialing cyberinfrastructure.
As predictive algorithms and decision-support tools are rapidly integrated into clinical and administrative workflows, there is an urgent need to train the healthcare workforce on AI safety, algorithmic bias, and clinical oversight. Traditional multiple-choice assessments are insufficient for evaluating the nuanced clinical reasoning required in the age of AI.
This repository serves as the central hub for crowdsourcing, developing, and peer-reviewing an open-access, role-specific curriculum designed to be deployed on our logic-based, AI-assisted credentialing platform.
Our cyberinfrastructure utilizes a node-based visual editor, localized Large Language Models (LLMs), and mandatory "rationale assessments" to evaluate a learner's underlying thought process when interacting with clinical AI.
However, the technology is only as good as the clinical expertise behind it. Our goal is to build a massive, globally searchable repository of high-fidelity, peer-reviewed clinical simulations spanning the entire hospital ecosystem.
We are issuing an open call to the global medical education community. We need subject matter experts to help us author branching scenarios for the following functional domains:
- Physicians & Medical Providers (Surgical, Medical, Intensive Care, Diagnostics)
- Advanced Practice & Midlevel Providers (NPs, PAs, CNMs)
- The Nursing Workforce (Inpatient, Outpatient, ICU, Informatics)
- Allied Health & Therapeutics (PT/OT, Respiratory, Pharmacy)
- Imaging, Radiology, & Laboratory Services
- Health Information, Billing, & Revenue Cycle
- Quality, Safety, Compliance, & Hospital Operations
There are currently 19 scenarios. Contributions are welcome — see How to Contribute below.
| Scenario | Topic | Focus | Decisions | Endings |
|---|---|---|---|---|
| Case Study: Early Deterioration and Conflicting AI Signals | unit1-recognizing-risks | Navigate a realistic clinical scenario involving patient Mr. Allen to learn about automation bias, over-reliance, and workflow integration. | 5 | 7 |
| The Drowning Heart | Heart Failure | acute and chronic management of Heart Failure | 6 | 4 |
| The 40-Week Clock | prenatal care | screening windows and preventative protocols | 4 | 3 |
| The Missing Milestone | primary amenorrhea | Evaluation and diagnosis | 7 | 7 |
| Time is Brain | ischemic stroke | acute management of an ischemic stroke | 4 | 5 |
| The Growing Pains | pediatric developmental milestones | Evaluation and parent education | 4 | 4 |
| The Golden Hourglass | neonatal jaundice | Diagnosis and management | 4 | 5 |
| The Breathless Traveler | pulmonology and emergency triage | Pulmonary Embolism (PE) | 5 | 5 |
| The Morning Ache | Rheumatoid Arthritis | Rheumatology | 5 | 5 |
| The 48-Hour Window | substance withdrawal | Psychiatry and Addiction Medicine | 5 | 5 |
| The Waddell Impact | Emergency care/Surgery | Multi-system trauma | 5 | 5 |
| The Peripheral Shadow | Lung cancer | Oncology | 5 | 5 |
| The Black Eschar | Sinusitis | ENT | 5 | 5 |
| The Restaurant Reaction | Anaphylaxis | Allergy and Immunology | 5 | 5 |
| The Winter Crisis | Sickle cell anemia | Hematology | 5 | 5 |
| The Daycare Diarrhea Dilemma | Diarrhea | Infectious disease/GI | 5 | 5 |
| The Vascular Toll | Diabetes Mellitus | Endocrinology | 5 | 5 |
| The Ring-Enhancing Dilemma | HIV | Infectious disease | 5 | 5 |
| The Gradient Collapse | Electrolytes and drug interaction | Pharmacology/Multi-systems | 5 | 5 |
scenarios/<scenario-name>/
├── <scenario-name>.json # the scenario graph: nodes, edges, viewport
├── README.md # the scenario told as a story, with a flowchart
└── images/ # images shown in the scenario
The JSON file is a graph of nodes connected by edges. Each node has a type:
| Node type | What it is |
|---|---|
informationNode |
A story page: what is happening to the patient. The first page has hideTopHandle: true. |
decisionNode |
A decision point where the learner chooses what to do. |
choiceNode |
One option at a decision, labelled A, B, C… and marked correct or misstep. |
terminationNode |
An ending: checkpoint (a concept learned) or fail (try again). |
You do not need to be a software developer to contribute! We are looking for clinical logic and scenario narratives.
The easiest way to contribute is by opening an Issue.
For accepted scenarios, we will collaborate in the discussion threads to map out the "Correct Pathways," "Recovery Loops" (where learners safely see the consequences of minor errors), and "Terminal Fail-States" (where severe clinical or safety breaches occur).
Help us peer-review existing scenario drafts to ensure they meet the highest standards of evidence-based practice and accurately reflect localized or international clinical guidelines.
The scenarios developed here will be integrated into a web-based cyberinfrastructure featuring:
- No-Code Visual Authoring: Transforming your text-based logic into interactive branching simulations.
- Local Sovereign AI Engines: Protecting institutional data and evaluating learner logic in real-time.
- Adaptive Delivery: Delivering your scenarios via classic text, interactive Socratic chat, or immersive visual environments.
The curriculum content generated in this repository is licensed under the GNU Affero General Public License v3.0.
If you use this framework or platform architecture in your research, please cite our upcoming publication:
Vald, G., Sermet, Y., & Demir, I. (Upcoming). Scalable Micro-Credentials for AI Literacy in Healthcare: An Open-Source Framework and Call for Collaborative Curriculum Development. International Journal of Educational Technology in Higher Education.
For institutional partnerships or questions regarding the cyberinfrastructure deployment, please reach out via GitHub Issues or contact the research team at Tulane University.
Core Team: Gabriel Vald, Yusuf Sermet, Ibrahim Demir (School of Science and Engineering, School of Medicine, Tulane University)