medimaging

medical imaging analysis using dicom server

Overall process interaction diagram

The above diagram has the following component:

nginx server to expose the orthanc server to the outside world.
Orthanc server which hosts the dicom protocol based REST api and the corresponding file storage on the server.
process cron, which regularly executes jobs to cleanup images which are imported from the orthanc server to the web viewer.
Web viewer backend REST api server that interacts with the orthanc server and does custom frontend based jobs
web ui: custom viewer for the images.

Portal images

Installation instructions

The software uses nodejs for running the backend REST API and Orthanc DICOM server to interface over REST API with dicom images. Hence you would require nodejs version > v10.16.3. Current version used to test is: v10.16.3. For unit testing we need to have mocha installed.

Steps:

- git clone https://github.com/nilaysaha/medimaging.git

Create binary for orthanc server

- cd server/Orthanc-1.7.2
- ./build_static.sh (This imports static files from remote repo and build a orthanc server in a newly created "Build" directory)
- nohup ./Build/Orthanc & (to start the dicom complaint REST api server and detach it from terminal)

Run a cron job for processing incoming images.

- node src/monitor_changes.js & (This runs a job every 10 seconds which checks the Orthanc server started above, for new images. If found it parses and converts images to png for frontend viewer)

Install the web viewer

- cd medimaging/src/viewer
- npm install (uses package.json to install dependencies and create node_modules directory)
- npm start (to start the web application for viewing the images)

Testing:

There is a unit test created to see if the Orthanc server is working.
To run the unit test:
- cd medimaging
- mocha tst/test.js (this tests the server is up and running, inputting a dicom image & deleting the dicom image)

License

MIT