Skip to content

Latest commit

 

History

History
135 lines (110 loc) · 3.46 KB

README.md

File metadata and controls

135 lines (110 loc) · 3.46 KB
Raw

pflink

Build

A Python-FastAPI application to create and manage workflows using PFDCM & CUBE .

User can register DICOMs to CUBEand additionally create new feed, add new plugin or pipeline in CUBE .

Quick Start

git clone https://github.com/FNNDSC/pflink.git
cd pflink
./pflink.sh

An instance of mongodb & pflink should be up and running on localhost on the following ports:

app URL
mongoDB http://localhost:27017
pflink http://localhost:8050

Go to http://localhost:8050/docs for API usage

Testing

cd pflink
./test.sh

Unmake

cd pflink
./unmake.sh

Examples

The POST API endpoint to create a new workflow or to get the status of an existing workflow is (/api/v1/workflow/)

sample workflow request

{
  "ignore_duplicate": true,
  "pfdcm_info": {
    "pfdcm_service": "PFDCM",
    "PACS_service": "orthanc"
  },
  "PACS_directive": {
    "StudyInstanceUID": "12365548",
    "SeriesInstanceUID": "66498598"
  },
  "workflow_info": {
    "feed_name": "test-%SeriesInstanceUID",
    "plugin_name": "pl-simpledsapp",
    "plugin_version": "2.1.0",
    "plugin_params": "--args ARGS"
  },
  "cube_user_info": {
    "username": "chris",
    "password": "chris1234"
  }
}

sample workflow response

{
  "status": true,
  "workflow_state": "initializing workflow",
  "state_progress": "0%",
  "feed_id": "",
  "feed_name": "",
  "message": "",
  "duplicates": null,
  "error": "",
  "workflow_progress_perc": 0
}

Additional support scripts

setup.sh

After starting new instances of pflink & mongo locally, we can a setup script available in the repo. The setup script provides the following support:

  1. Authentication (authenticate into any running instance of pflink)
  2. Add a new PFDCM service to MongoDB

to use:

cd pflink/scripts
./setup.sh --help

resetWorkflow.sh

Sometimes, we need to find a specific workflow request and restart using pflink. This repo contains a script resetWorkflow to do so. The script does the following tasks:

  1. Authenticate to a pflink instance
  2. Find a list of "keys" pointing to workflow records that match the user search values
  3. Display each workflow record and provide the following options:
    1. delete the record and re-submit
    2. continue to the next record
    3. exit

to use:

cd pflink/scripts
./resetWorkflow.sh --help

dateSearch.sh

This is one of the most important script to use. We can use this script for find the total number of studies present for a given study name in a give date range, and verify if any series in those studies are processed by pflink and pushed to SYNAPSERESEARCH.

We can also use this script to analyze all the studies matching a study name in a given date range. The parameters we need to provide to run this script is a start date, an end date, a keyword for study name and the mode we want to run this script on:

  1. search
  2. analyze

to search:

cd pflink/scripts
./dateSearch.sh -S 2024-07-03 -E 2024-07-08 -K "XR HIPS TO ANKLES LEG MEASUREMENTS" -D search

to analyze:

cd pflink/scripts
./dateSearch.sh -S 2024-07-03 -E 2024-07-08 -K "XR HIPS TO ANKLES LEG MEASUREMENTS" -D analyze