AWS HealthLake provides integrated Natural Language Processing (NLP) libraries that parse, identify, and map information from unstructured data stored in FHIR DocumentReference resources. This process identifies entities with traits like SIGN, SYMPTOM, or DIAGNOSIS, creating new Condition and Observation resources linked to the source document.
AWS HealthLake allows users to export FHIR data. Users can use native actions to start and manage export jobs, which are processed asynchronously in a FIFO (First In First Out) manner.
AWS HealthLake allows the management of various FHIR resources through RESTful APIs compliant with the HL7 FHIR R4 standard. Developers can utilize specific interactions (create, read, update, etc.) to achieve efficient data exchange within the data store.
AWS HealthLake supports the base FHIR R4 specification and validates data using various domain-specific 'FHIR Profiles.' This ensures that data imported into the service adheres to diverse international standards (such as US Core, UK Core, etc.), enabling robust handling of specialized medical information.
This article provides comprehensive guidance on building a full-scale FHIR system using Amazon Web Services (AWS). It focuses on key architectural pillars, including security, reliability, and performance efficiency.