MMedIngest
Platform
Solutions
Industries
Developers
Pricing
Resources
ConsoleBook Demo
← Back to all articles
Interoperability

Mapping Scanned Referrals to FHIR Observations

Step-by-step tutorial on translating raw OCR document text segments into fully compliant, queryable US Core FHIR resources.

By Sarah JenkinsJuly 12, 20261 min read

Table of Contents

Mapping Text to CodesConstructing the Observation Resource

Interoperability in modern healthcare relies on HL7 FHIR standards. This tutorial outlines how to compile parsed OCR parameters into queryable Observation bundles.

Mapping Text to Codes

Before writing to FHIR, raw text values must match standardized vocabularies:

  • Medications -> **RxNorm**
  • Lab Results -> **LOINC**
  • Diagnoses -> **SNOMED CT**

Constructing the Observation Resource

A typical blood sodium observation resource is compiled as follows:

snippet.json
{
  "resourceType": "Observation",
  "status": "final",
  "code": {
    "coding": [
      {
        "system": "http://loinc.org",
        "code": "2951-2",
        "display": "Sodium [Serum/Plasma]"
      }
    ]
  },
  "valueQuantity": {
    "value": 138,
    "unit": "mmol/L"
  }
}

This resource is then bundled and written to the EHR via SMART on FHIR write APIs.

Share this article:
MMedIngest

AI-native clinical document intelligence automating ingestion, mapping standard terminologies, and validating FHIR.

Subscribe to our newsletter

Product

  • Platform Overview
  • OCR Pipeline
  • Medical AI Extraction
  • Terminology Mapping
  • FHIR Generation API
  • Pricing Tiers

Solutions

  • Hospitals & Systems
  • Diagnostic Laboratories
  • Clinics & Groups
  • Revenue Cycle (RCM)

Resources

  • Security & Compliance
  • API Sandbox
  • Developer Docs
  • Case Studies & ROI
  • Research Blog

Company

  • About MedIngest
  • Careers
  • Press & Brand Assets
  • Contact Sales

© 2026 MedIngest, Inc. All rights reserved. HIPAA Compliant. SOC 2 Type II Certified.