qiicr / dcm2tables

Overview

Tools to go from a DICOM dataset into a collection of tables for the purposes of data repurposing and exploration using relational databases.

This is not a general purpose tool. Initial goal is to support exploration of the TCIA QIN-HEADNECK dataset.

Prerequisites

Software:

• python
• pandas
• pydicom (must be installed from source! distribution available via pip is too old and does not have some of the DICOM dictionary entries we parse)
• dcmqi is required to generate certain object-specific tables for such objects as DICOM Segmentations and measurements structured reports; it is best if you make sure dcmqi converters are in your path prior to running the script

Data:

• https://wiki.cancerimagingarchive.net/display/Public/QIN-HEADNECK

Usage

python tabulate.py -s schema.qdbd -d <location of a directory with DICOM data> -o <output location to store csv tables>

The script will create tables stored in CSV in the directory where it was executed.

Contact

• Andrey Fedorov https://fedorov.github.io

dcm2tables's Issues

Add pointers to the schema

Consider how to handle private attributes

In some situations it might be helpful to be able to handle private attributes without changing the code.

Add Dockerfile

Make it possible to use the tool from a docker image

Add progress reporting