Managing and Viewing DICOM Images with ImageJ
Johannes Hermen
a
, Christian Moll
a
, Andreas Jahnen
a
a
Public Research Centre Henri Tudor - CR SANTEC, 2A, rue Kalchesbreck, L-1852,
Luxembourg;
ABSTRACT
ImageJ is capable to read DICOM (Digital Imaging and Communication in Medicine)
3
images, the standard
format in medical imaging. But this is not enough: to receive, to send, to compress, to decode, to compare or to
search you need dedicated plugins. This paper reviews the ImageJ features related to DICOM and present the
different existing ImageJ plugins that help to get the problem solved with ImageJ. The demonstration includes
information on how to perform the most important tasks within DICOM Imagery and which commands and
plugins to use. The weaknesses and limitations of ImageJ are discussed as well. The features of the TUDOR
DICOM Tools, a collection of DICOM Plugins for ImageJ are shown including a presentation of the included
DICOM Viewer.
Keywords: ImageJ, DICOM, medical image processing
1. INTRODUCTION
Over the last years the classical, analog medical imaging modalities are more and more replaced by digital image
acquisition technologies. Today Computers have taken an important part in the diagnostic medical imaging.
The mostly used format for storing, transferring and processing these digitalized medical images is the DICOM
(Digital Imaging and Communication in Medicine) standard. The DICOM standard defines an image format as
well as formats for sending, receiving and querying DICOM object (e.g. images) via network. The image format
which does not only contain the pixel data of the acquired image, it also contains meta data about the patient,
the used equipment and the acquisition parameters. This data is stored in tags in the so-called DICOM Header.
Each tag contains a key in the form “0002,0010” and a corresponding value. When working with medical images,
these Meta-data is nearly as important as the image itself to process, store and evaluate the image. Because of
this, all software-solutions working with DICOM images, do allthough have to care about these stored meta-data.
ImageJ, a free Java based image processing software and library is capable of reading DICOM images and
their header data.
This paper will give an overview of the built-in functionalities of ImageJ for working with medical images
and how they can be extended by free available plugins. Afterwords we will show the TUDOR DICOM Tools, a
set of Java libraries developed by our team containing ImageJ plugins and stand-alone applications for working
with DICOM images. These tools combine a lot of needed features to an easy to use set of plugins or a base for
own custom DICOM applications.
2. IMAGEJ BUILT IN FUNCTIONALITIES
ImageJ itself is capable of opening single DICOM images or image stacks from the local harddisk. It uses a
build-in plugin, implemented in the ij.plugin.DICOM class to provide these features. Just like any other image
format DICOM images can be loaded into ImageJ by using the open command (File ⇒ Open) or just dropping
the images onto the application. Image stacks can be loaded by using the import image sequence command
(File ⇒ Import ⇒ Image Sequence). After choosing one of the stack images in a folder, ImageJ will import all
images in this folder into one image stack. During the import you can filter these images by image name or
image number and choose the starting image as well as the number of images to import. The meta-data of the
images, the so-called DICOM-Header is written to the ImagePlus object and can be accessed via the show info
command (Image ⇒ Show.Info)
This built-in plugin is only capable of importing DICOM files with raw, uncompressed image data. Images
with JPEG
4
or Lossless Run Length Encoding (RLE) can not be openend. In this case ImageJ will present an