Derma DICOMiser

Python tool to generate test DICOMS that follow the BAD AI Dermatology standard

Why was this tool built?

This tool was build to allow the generation of DICOM data for use as part of the MELODY project.

What does this tool do?

The dicomiser takes a metadata CSV and any number of images (JPGs, PNGs etc) and generates a single multi-frame DICOM.

Prerequisites

• Python3
• DCMTK

Setup

2. Set an environment variable called img2dcm that points to the img2dcm DCMTK tool e.g. $img2dcm=~/jfriel/dmctk/bin/img2dcm
3. install dependencies pip install -r requirements.txt

Usage

python dicomiser.py OUTPUT_DIRECTORY METADATA_FILE IMAGE...

Metadata file

A template metadata CSV is provided within this repository. It contains entries for all values that may be populated within the generated DICOM. All values are optional. Test Data can be generated by running the testDataGenerator.py script

Limitations

• Multi-Frame images are currently very restricted, and require all images to be the same type and size
• Currently only works with JPGs