You are an expert healthcare AI governance analyst specializing in Canadian medical device regulations and health data privacy laws. Generate a comprehensive Quality Metrics Report for the AI system described below.

**System Context:**
- AI System Name & Version: [AI_SYSTEM_NAME]
- Healthcare Domain: [HEALTHCARE_DOMAIN] (e.g., diagnostic imaging, clinical decision support, pathology)
- Deployment Setting: [DEPLOYMENT_CONTEXT] (e.g., tertiary care hospital, community health centre, rural clinic)
- Canadian Province/Territory: [PROVINCE] (for specific privacy law compliance)
- Evaluation Period: [EVALUATION_PERIOD]
- Health Canada Regulatory Class: [REGULATORY_CLASS] (Class I, II, III, or IV under Medical Devices Regulations)
- Dataset Provenance: [DATA_SOURCES] (training/validation data origins)

**Report Requirements:**

1. **Executive Dashboard**
   - Overall Quality Score (0-100) with 95% confidence intervals
   - Risk classification justification per Health Canada SaMD guidance
   - Critical alerts requiring immediate regulatory or clinical attention
   - Compliance status summary (Red/Yellow/Green) for each major category

2. **Clinical Performance Metrics**
   - Diagnostic accuracy (sensitivity, specificity, PPV, NPV) against Canadian population benchmarks
   - Temporal drift analysis (performance stability over [EVALUATION_PERIOD])
   - Subgroup performance analysis including Indigenous populations, rural vs. urban, linguistic minorities, and age demographics
   - Comparison against current standard of care in Canadian healthcare settings

3. **Equity & Bias Assessment**
   - Demographic parity metrics across protected groups under Canadian Human Rights Act
   - Socioeconomic bias detection (income quintile, postal code analysis)
   - Accessibility compliance (WCAG 2.1 AA standards for interface components)
   - Health equity impact assessment specific to Canadian healthcare disparities

4. **Regulatory Compliance Matrix**
   - Health Canada SaMD pre-market/submission requirements checklist
   - Provincial privacy law adherence (PHIPA, PIPEDA, or applicable provincial legislation)
   - Federal Algorithmic Impact Assessment (AIA) compliance level and requirements
   - ISO 13485 (Quality Management) and ISO 14971 (Risk Management) alignment status
   - Data residency compliance (Canadian privacy law requirements)

5. **Safety & Reliability Indicators**
   - Failure mode analysis adapted for Canadian healthcare workflows
   - Adverse event simulation and mitigation strategies
   - Human-AI interaction safety metrics (override rates, automation bias indicators)
   - Cybersecurity resilience score (HIPAA/PIPEDA security safeguards)

6. **Data Governance & Provenance**
   - Training data representativeness (Canadian demographic coverage analysis)
   - Data lineage documentation and version control status
   - Consent and authorization compliance for Canadian health data
   - Retention and disposal policy alignment with provincial health authority guidelines

7. **Interpretability & Transparency**
   - Clinical explainability scores (LIME/SHAP metrics for relevant predictions)
   - Decision audit trail completeness
   - Clinician-facing explanation quality assessment
   - Patient communication material readability (Flesch-Kincaid Grade Level)

8. **Actionable Recommendations**
   - Specific mitigation strategies for identified risks with timelines
   - Re-validation and re-certification schedule per Health Canada requirements
   - Provincial implementation considerations (interoperability with existing EHR systems like Epic, Meditech, or provincial health networks)
   - Resource requirements for maintaining compliance

**Formatting Standards:**
- Use Canadian English spelling (behaviour, centre, colour, analyse)
- Reference specific sections of the Medical Devices Regulations (SOR/98-282) where applicable
- Include confidence intervals for all statistical measures
- Provide metric thresholds based on Health Canada SaMD guidance and provincial health authority standards
- Structure for executive review with technical appendices suitable for regulatory submission