dcmqi: A hands-on introduction

use the left and right arrow keys to navigate the slides topics

use the down and up arrow keys to drill down into the topic details

Andrey Fedorov, Brigham and Women's Hospital

http://qiicr.org

Contents

Motivation

  • DICOM is the format of images that we get from clinical equipment
  • Quantitative imaging is the extraction of quantifiable features from medical images
  • Images are pixels + metadata in the DICOM-defined format
  • Metadata is typically lost while converting from DICOM into research formats
  • Metadata is most often missing in the results produced by the image analysis tools

dcmqi = DICOM for Quantitative Imaging

  • conversion of common derived image data types produced by quantitative image analysis
    • segmentations, parametric maps, scalar measurements over segmented regions of interest
  • command line converters
  • C++ Application Programming Interface
  • schemas for metadata

This tutorial will teach you how to ...

  • set up your computer to use dcmqi
  • convert basic segmentation into DICOM
  • visualize DICOM images and segmentations

Video 1: Setup

What you will need to repeat the steps in the video

  • Docker
  • 3D Slicer
  • Atom
  • If you are using Windows, we recommend PowerShell for running Docker commands

What we tested

  • Tutorial was recorded on a Mac OS X 10.10.5 Yosemite
  • We also tested Docker and 3D Slicer operation on Windows 10 Pro and Ubuntu 14.04
  • Docker requires Windows 10 Pro 64-bit or above!
  • Configuration of the Atom editor to enable dicom-dump on Linux and Windows is a bit different - please see these instructions

Video 2: First conversion task

Note: the tutorial was recorded in a Mac environment! If you are using Windows, your command line format will have to be different! See our documentation for details!

What you will need to repeat the steps in the video

Video 3: Visualizing the result

What you will need to repeat the steps in the video

  • TCIA collection TCIA-HEADNECK subject 139 (don't forget to unzip the top-level package and all the compressed files underneath!)
  • You can also download this dataset using TCIA Browser module of 3D Slicer that we installed in the Part 1 of the tutorial

Want to know more?