0.1.0 - ci-build
TrustAITransparencyIG - Local Development build (v0.1.0) built by the FHIR (HL7® FHIR® Standard) Build Tools. See the Directory of published versions
| Use Case | Title |
|---|---|
| Use Case 1 | AI-supported Risk evaluation |
| Use Case 2 | AI-generated Diagnostic Report |
The example instances of this IG are grouped into two example sets. All example resources are listed on the Artifacts page.
| Example Set | AI System | AI Output | Related Use Case |
|---|---|---|---|
| Specialized AI Output Scenario | RiskAssist AI | Trust_AIObservation | Use Case 1 |
| Generalized AI Output Scenario | DiagnosticAssist AI | DiagnosticReport | Use Case 2 |
The specialized example set contains four scenarios based on the same synthetic patient and AI system:
| Scenario | Description | AI Output |
|---|---|---|
sc-01-ai-only |
Core AI execution and traceability only | sc-01-ai-only-ai-observation-risk-001 |
sc-02-validation |
AI output accepted by human reviewer | sc-02-validation-ai-observation-risk-001 |
sc-03-override |
AI output overridden by human reviewer | sc-03-override-ai-observation-risk-001 |
sc-04-correction-exp |
AI output corrected and explained to patient | sc-04-correction-exp-ai-observation-risk-001 |
The generalized example set shows how the IG can be used for AI outputs without a dedicated profile. DiagnosticAssist AI generates a draft DiagnosticReport from a lab result, which is validated by a human reviewer before it is shared with the patient. See Use Case 2.