protoncluster / Niffler

A DICOM Framework for Machine Learning Pipelines against Real-Time Radiology Images

Niffler: A DICOM Framework for Machine Learning Pipelines against Real-Time Radiology Images

Niffler is a real-time PACS receiver and metadata extractor framework. It consists of a metadata extractor at its core. MetadataExtractor.py in src/meta-extraction implements the core extraction functionality.

Using Niffler

Make sure to configure the PACS to send data to Niffler's host, port, and AE_Title. Niffler won't receive data unless the PACS allows the requests from Niffler (host/port/AE_Title).

In the src folder:

• cold-extraction: On-demand queries to retrieve retrospective DICOM data.

• meta-extraction: Extracts metadata from a continuous real-time DICOM imaging stream.

• png-extraction: Converts a set of DICOM images into png images, extract metadata in a privacy-preserving manner.

• scripts: The scripts to enable ML Algorithm containers.

• main: The Application Layer, developed in Java

Developing Niffler

Niffler core (cold-extraction, meta-extraction, and png-extraction) is built with Python3. The scanner utilization tool in the Application layer is built in Java and the scripts are developed in Javascript.

To develop the Niffler core, first, install the dependencies.

$ pip install requests pymongo schedule pydicom pynetdicom

For the development branch of pynetdicom

$ pip install git+git://github.com/pydicom/pynetdicom.git

Also install DCM4CHE from https://github.com/dcm4che/dcm4che/releases

For example,

$ wget https://sourceforge.net/projects/dcm4che/files/dcm4che3/5.22.5/dcm4che-5.22.5-bin.zip/download -O dcm4che-5.22.5-bin.zip

$ sudo apt install unzip

$ unzip dcm4che-5.22.5-bin.zip

Make sure Java is also available, as DCM4CHE requires Java to run.

Also, the scanner utilization processing of Niffler is developed in Java.

Then checkout Niffler source code.

$ git clone https://github.com/Emory-HITI/Niffler.git

$ cd Niffler

The master branch is stable whereas the dev branch has the bleeding edge.

The Java components of Niffler Application Layer are managed via Apache Maven 3.

Please refer to each module's individual README for additional instructions.

Citing Niffler

If you use Niffler in your research, please cite the below paper:

• Pradeeban Kathiravelu, Puneet Sharma, Ashish Sharma, Imon Banerjee, Hari Trivedi, Saptarshi Purkayastha, Priyanshu Sinha, Alexandre Cadrin-Chenevert, Nabile Safdar, Judy Wawira Gichoya. A DICOM Framework for Machine Learning Pipelines against Real-Time Radiology Images. arXiv preprint arXiv:2004.07965 (2020).