Healthcare lakehouse
The foundational element of healthcare data solutions in Microsoft Fabric is the healthcare lakehouse. The healthcare lakehouse is built on a modern data architecture that applies Fast Healthcare Interoperability Resources (FHIR) and Digital Imaging and Communications in Medicine (DICOM). The low-cost, scalability, and flexibility healthcare lakehouse supports structured clinical, unstructured clinical, and medical imaging data.
Healthcare lakehouse is developed on medallion lakehouse architecture, which is purpose-built for healthcare customers. The goal is to improve the structure and quality of data incrementally and progressively as it flows through each layer of the architecture (from bronze to silver to gold layer tables).
The following diagram shows levels in medallion lakehouse architecture and the supported models that each level uses.
The healthcare data solutions medallion architecture has three fundamental layers:
Bronze - Also called the raw zone, this first layer organizes the source data in its original format. The bronze layer also stores the structured and semi-structured content of the source data in delta tables, also referred to as staging tables. These tables are compressed and indexed for more efficient transformation processing. The data in this layer is typically append-only and immutable. The staging tables are considered part of the internal implementation, and you shouldn't modify or extend them to customize your deployment.
Silver - Also called the enriched zone, this layer stores data that's sourced from the bronze layer. In the silver layer, data undergoes refinement processes, including validation checks and enrichment techniques, to enhance its accuracy and utility for downstream analytics. Data in the silver lakehouse has certain rules in place to determine when to update an existing record or insert a new one. These rules are based on the deterministic ID and the last modified date and time of each record.
Gold - Also called the curated zone, this final layer stores data that's sourced from the silver layer. The data is refined to meet specific downstream business and analytics requirements. This layer serves as the primary source for high-quality, aggregated datasets that are ready for comprehensive analysis and insights extraction.
In addition to the lakehouses that support the core data architecture, the Admin lakehouse is also available. The Admin lakehouse holds all files and tables that provide governance, traceability, and management across all other medallion lakehouses, such as the global configuration and validation errors in the BusinessEvent table.