HSE CTS Flask Explorer

A local Flask web application for exploring the HSE Central Terminology Service (CTS) FHIR API. The project provides a secure, practical interface for searching terminology content, validating codes, expanding ValueSets, and inspecting FHIR responses without exposing CTS credentials in the browser.

Purpose

Healthcare developers and informatics users often need a safe way to inspect terminology services before wiring them into applications. This project wraps the HSE CTS API with a server-side Flask application so users can explore SNOMED CT, LOINC, UCUM, CodeSystems, and ValueSets while keeping OAuth2 credentials local and server-side.

Core Workflows

• Search SNOMED CT concepts by clinical term
• Look up exact SNOMED CT, LOINC, and UCUM codes
• Browse available CodeSystems and ValueSets
• Expand ValueSets through CTS
• Search UCUM units directly by unit text, such as mercury or kg/m2
• Find likely UCUM units for clinical observations using LOINC EXAMPLE_UCUM_UNITS

Technical Implementation

The application follows a server-side proxy pattern:

• Browser requests go to Flask
• Flask handles OAuth2 client-credentials authentication
• Authenticated FHIR requests are sent from the server to CTS
• Jinja templates render searchable, readable terminology results
• Raw FHIR JSON remains available for debugging and implementation reference

The local structure separates API authentication, FHIR client logic, operation helpers, route handlers, templates, static assets, and documentation.

Terminology Features

The project supports practical terminology exploration workflows including exact code lookup, direct UCUM search, CodeSystem browsing, ValueSet browsing, and LOINC-backed observation-to-unit discovery.

For example, a search for a clinical observation such as “body mass index” can locate the relevant LOINC observation, inspect its EXAMPLE_UCUM_UNITS, and show likely UCUM units such as kg/m2.

Security And Use Constraints

The project is designed for local development and terminology exploration:

• CTS credentials stay in .env
• Client secrets are never sent to the browser
• OAuth2 access tokens are cached and refreshed server-side
• The expected CTS account is read-only
• The repository should not contain live credentials, tokens, or exported terminology datasets
• The application is not a clinical system

What This Demonstrates

This project demonstrates secure terminology-service access, FHIR operation design, OAuth2 proxy architecture, practical developer tooling, and the ability to turn standards infrastructure into something easier to inspect, test, and explain.