DICOM Helper Scripts

This repository contains a collection of bash scripts designed to streamline various tasks related to handling and processing DICOM (Digital Imaging and Communications in Medicine) data. Some scripts leverage Google Cloud services, such as the Healthcare API for interacting with DICOM stores.

Scripts

General DICOM Utilities

• ccli.sh: Provides a convenient command-line interface for executing common DICOM utilities within a Docker container. Supports both dcmtk and dcm4che toolsets.

Study Creation Tools

• create_study/makestudy.sh: Generates a series of DICOM files with varying study and instance numbers, useful for testing and populating DICOM stores.

• create_study/gen_study/gen_study.py: Python script for generating DICOM study data with configurable parameters.

DIMSE Operations

• dimse/findstudies.sh: Search a PACS and create a JSON file with metadata for all the responses

Google Cloud Healthcare API Integration

De-identification

• healthcareapi/deid/deid_dicom_store.sh: De-identify a set of DICOM files in a DICOM store using Google Cloud Healthcare API

Data Insertion

• healthcareapi/insert/insert_file.sh: Uploads a single DICOM file to a specified DICOM store using the Healthcare API v1beta1 endpoint via PUT method. Includes comprehensive help documentation accessible with --help.

• healthcareapi/insert/insert_folder.sh: Uploads all DICOM files within a given folder to a DICOM store using the Healthcare API v1 endpoint via POST method. Supports parallel processing, selective study deletion, and single-study mode. Includes comprehensive help documentation accessible with --help.

• healthcareapi/insert/post.sh: A helper script that handles HTTP POST requests for DICOMweb interactions using the Healthcare API v1 endpoint. Constructs its own API endpoint and handles rate limiting with automatic retries.

• healthcareapi/insert/put.sh: A helper script that handles HTTP PUT requests for DICOMweb interactions using the Healthcare API v1beta1 endpoint. Constructs its own API endpoint and handles rate limiting with automatic retries.

Data Retrieval

• healthcareapi/retrieve/get_instance_list.sh: Retrieve a list of DICOM instances from a Healthcare API DICOM store

• healthcareapi/retrieve/perf_test.sh: Performance testing script for Healthcare API operations with detailed metrics and output logging

Prerequisites

• General:

  • Bash shell
  • Docker (for ccli.sh)
  • dcmodify (for create_study/makestudy.sh)
  • Python 3 with required packages (for create_study/gen_study/gen_study.py)
  • findscu and dcm2json (for dimse/findstudies.sh)

• Healthcare API Scripts:

  • Google Cloud Project with Healthcare API enabled
  • Service account with appropriate permissions
  • gcloud command-line tool
  • curl and jq (for JSON processing)
  • dcm2bq (for healthcareapi/insert/insert_folder.sh)

Usage

Getting Help

Most scripts include built-in help documentation. Use the --help flag for detailed usage information:

# Get help for Healthcare API scripts
./healthcareapi/insert/insert_file.sh --help
./healthcareapi/insert/insert_folder.sh --help

Healthcare API Scripts

Single File Upload:

./healthcareapi/insert/insert_file.sh \
  projects/PROJECT_ID/locations/LOCATION/datasets/DATASET_ID/dicomStores/DICOM_STORE_ID \
  path/to/file.dcm

Folder Upload (Basic):

./healthcareapi/insert/insert_folder.sh \
  projects/PROJECT_ID/locations/LOCATION/datasets/DATASET_ID/dicomStores/DICOM_STORE_ID \
  /path/to/dicom/folder

Folder Upload (Advanced Options):

# Single study, skip deletion, 20 parallel processes
./healthcareapi/insert/insert_folder.sh -s -d -p 20 \
  projects/PROJECT_ID/locations/LOCATION/datasets/DATASET_ID/dicomStores/DICOM_STORE_ID \
  /path/to/dicom/folder

Refer to the individual script files and their --help output for detailed usage instructions and examples.

Contributing

Contributions are welcome! Feel free to submit pull requests for bug fixes, new features, or improvements to existing scripts.

License

This project is licensed under the Apache 2.0 License - see the LICENSE file for details.