home / mcp / taiwan icd10 health mcp server
Provides integrated Taiwan ICD-10, FDA products, health foods, nutrition data, LOINC, and clinical guidelines via FHIR R4.
Configuration
View docs{
"mcpServers": {
"annie25726-taiwan-health-mcp-main": {
"command": "python",
"args": [
"src/server.py"
]
}
}
}You can access a comprehensive, Taiwan-localized MCP server that integrates ICD-10 diagnoses, FDA drug data, health foods, nutrition, LOINC laboratory codes, and clinical guidelines using the FHIR R4 standard. This MCP server enables standardized querying, data conversion to FHIR resources, offline deployment options, and seamless integration with MCP clients for clinical decision support and research.
You will interact with the MCP server through an MCP client that communicates using the MCP protocol. Use the server to query ICD-10 diagnoses and procedures, retrieve and translate FDA drug data into FHIR Medication resources, analyze health foods and nutrition data, access LOINC lab codes and reference values, and explore clinical guidelines and pathways. You can combine these data sources to build clinical decision support workflows, perform data conversions to FHIR resources, and run offline analyses when network access is limited.
# 1. Clone the project
git clone https://github.com/audi0417/Taiwan-Health-MCP.git
cd Taiwan-Health-MCP
# 2. Install dependencies
pip install -r requirements.txt
# 3. Prepare data (download ICD-10 Excel file into the data/ directory)
# 4. Start the service
python src/server.py
```
```bash
# Docker-based quick start (recommended for isolation)
docker-compose up -dOffline deployment is supported. If you must run without internet access, first initialize the data in a connected environment to generate database files, then move them to the offline environment and enable offline mode by setting MCP_OFFLINE=true. Note that offline mode disables data updates and online downloads.
This server brings together multiple official Taiwan data sources and international standards, with planned ongoing updates. The health foods and dietary recommendations features are in development and require professional validation before clinical use. For offline or restricted environments, you should ensure all required data files (like ICD-10 Excel and databases) are present in the local data directory before starting the server.
The offline mode relies on locally stored data files. When enabled, automatic data synchronization is paused. Ensure you have the necessary data files ready in the data/ directory, including ICD-10 data and local copies of the drug, health food, and nutrition databases.
Query ICD-10-CM diagnoses and ICD-10-PCS procedure codes, infer complications, detect code conflicts, and convert results into FHIR Condition resources.
Aggregate five official FDA datasets (licenses, appearances, ingredients, ATC taxonomy, and labeling) and convert to FHIR Medication and MedicationKnowledge resources.
Query Taiwan FDA-approved health foods and health claims and analyze associations with diseases.
Access nutrition components, perform diet analysis, and query ingredients and additives.
Map common LOINC test codes, lookup reference ranges by age and sex, and interpret test results.
Query clinical guidelines from Taiwanese medical associations and outline diagnostic and treatment recommendations.
Transform data into FHIR Condition, Medication, and MedicationKnowledge resources to enable standard interoperability.
Provide steps to initialize local data and run in offline mode to support air-gapped environments.
Combine disease data with health and nutrition information to enable comprehensive analyses.