DICOM data wrangling

In this part of the tutorial we will learn how to work with a DICOM dataset spanning different TCIA collections and containing various types of DICOM objects.

Hands-on demo notes

• TCIA as a use case

  • LIDC-IDRI (annotations in XML)

  • TCGA-GBM (annotations in NIfTI)

  • TCGA-LGG (annotations in NIfTI)

  • QIN-HEADNECK

  • QIN-PROSTATE-Repeatability (not yet released)

• discuss the example dataset used in the demo

• steps for handling DICOM data:

  • dicomsort

  • dcm2niix for converting image series into volumes

  • dcmqi for working with SEG and SR

  • dcm2tables: conversion into tabular representation for working with metadata

    • "database" visual schema

  • Jupyter Notebook demonstration

This part will be covered in this Jupyter Notebook.

QIN-HEADNECK collection exploration

See this Jupyter Notebook we developed for DICOM4MICCAI 2017.