This module develops the evaluative skills needed to audit dataset lineage and access-control decisions for multimodal medical AI projects in Microsoft Foundry (formerly Azure AI Studio), verifying that inputs, transformations, and usage rights are correctly documented and traceable. Learners will be able to apply a structured lineage and access audit framework to a provided scenario, identify gaps or policy violations, and produce a documented audit summary. Lineage and access auditability are the cornerstones of responsible healthcare AI governance: without them, organizations cannot demonstrate compliance, trace data provenance, or identify the source of model reliability issues after deployment.

This module covers how to apply de-identification methods and minimum-necessary data practices in Microsoft Foundry (formerly Azure AI Studio) when preparing multimodal healthcare datasets—including imaging metadata, radiology report text, and patient-linked attributes—for bounded AI development or evaluation tasks. Learners will be able to apply a structured de-identification and minimum-necessary review approach to a provided multimodal scenario dataset, correctly documenting the de-identification method used and the minimum-necessary justification for each included data element. Applying these practices is the primary safeguard against patient privacy harm in healthcare AI development workflows and a prerequisite for HIPAA-aligned data use.

This module develops the evaluative skills needed to assess privacy risks in Microsoft Foundry (formerly Azure AI Studio) multimodal workflows, specifically identifying re-identification exposure across image metadata, report text, and linked patient attributes, and flagging required safeguards before workflow use. Learners will be able to apply a structured privacy risk assessment framework to a provided multimodal workflow scenario, identify re-identification risk factors across all three data dimensions, and produce a documented privacy risk assessment with safeguard recommendations. Privacy risk evaluation before workflow use is the critical quality gate that prevents re-identification harm, policy violations, and regulatory exposure in healthcare AI projects.

This module covers how to configure a Microsoft Foundry project for a medical imaging and multimodal AI workload—setting up the workspace, establishing model access, connecting data sources, and configuring evaluation flows within defined organizational constraints. Learners will be able to review a prepared configuration scenario and document the key configuration decisions made at each stage of the setup. Understanding project configuration is the operational foundation that determines whether a healthcare AI project is reproducible, auditable, and correctly scoped for its intended use case.

This module develops the analytical and evaluative skills needed to assess deployment pathway options and monitoring cost-performance tradeoffs for multimodal healthcare AI workloads in Microsoft Foundry . Learners will be able to distinguish development, testing, and controlled production deployment pathways, and evaluate whether configured monitoring resources and compute settings support reliable inference and review at the expected operational scale for a bounded medical imaging AI workload. These skills are essential for healthcare AI professionals responsible for deployment readiness decisions and operational sustainability assessments.

In this project, learners produce a governed multimodal AI project design, a professional artifact that integrates data governance controls, patient privacy protections, and Microsoft Foundry (formerly Azure AI Studio) project configuration and monitoring skills for a single realistic healthcare AI scenario. Learners will be able to apply three skills to evaluate, document, and recommend a complete governance, privacy, and operational configuration approach for a bounded Microsoft Foundry (formerly Azure AI Studio) multimodal medical AI workflow. Producing this kind of integrated governance design document is a core professional output for healthcare AI administration roles, where responsibility for responsible AI deployment spans governance stewardship, privacy compliance, and operational readiness.