Module 6 Assignment: Validation, bias, and safety

Scenario

You are advising a radiology service deciding whether an imaging AI should advance from retrospective review to supervised pilot use. The stakeholders are: radiologist, imaging technician, compliance officer, and patient safety lead.

Task

Answer the module question: What evidence is needed before clinical use?

Use the module lab and course readings to produce: imaging model evidence packet with preprocessing notes, validation limits, and clinical handoff risks focused on validation, bias, and safety: Create a validation plan with subgroup analysis..

Required Evidence

• Define the decision or system boundary in one paragraph.

• Identify the dataset, proxy data, or evidence source you used: synthetic 8x8 grayscale image arrays with a small bright lesion pattern and non-lesion variation.

• Compare at least two alternatives, baselines, policies, or designs.

• Report one quantitative result or structured scoring table.

• Explain two failure modes and one mitigation for each.

• State what additional evidence would be required before real deployment.

Submission

Submit the completed notebook plus a 900-1200 word memo. The memo must include clear headings for context, method, evidence, risks, recommendation, and open questions.

# Assignment workspace for Module 6: Validation, bias, and safety
module = 6
decision = "What evidence is needed before clinical use?"
artifact = "imaging model evidence packet with preprocessing notes, validation limits, and clinical handoff risks focused on validation, bias, and safety: Create a validation plan with subgroup analysis."

alternatives = [
    {"option": "baseline_or_manual_process", "strength": "", "risk": "", "evidence": ""},
    {"option": "ai_assisted_or_advanced_option", "strength": "", "risk": "", "evidence": ""},
]

recommendation = {
    "decision": decision,
    "recommended_option": "",
    "minimum_evidence_before_pilot": [],
    "monitoring_metric": "",
    "rollback_trigger": "",
}

{"module": module, "artifact": artifact, "alternatives": alternatives, "recommendation": recommendation}

{'module': 6,
 'artifact': 'imaging model evidence packet with preprocessing notes, validation limits, and clinical handoff risks focused on validation, bias, and safety: Create a validation plan with subgroup analysis.',
 'alternatives': [{'option': 'baseline_or_manual_process',
   'strength': '',
   'risk': '',
   'evidence': ''},
  {'option': 'ai_assisted_or_advanced_option',
   'strength': '',
   'risk': '',
   'evidence': ''}],
 'recommendation': {'decision': 'What evidence is needed before clinical use?',
  'recommended_option': '',
  'minimum_evidence_before_pilot': [],
  'monitoring_metric': '',
  'rollback_trigger': ''}}

Acceptance Criteria

Your submission is complete only if another reviewer can reproduce your reasoning from the evidence you provide. You do not need production-grade data, but you must be explicit about proxy-data limits and what would change with real institutional data.