From Bangor Imaging Unit
Jump to: navigation, search

3D Slicer is a image visualisation and analysis package. Slicer is multi-modal and multi-organ. MRI, CT, microscopy, brains to kidneys, Slicer has modules for all of them. Slicer does image registration, DTI and volume rendering. Slicer reads and writes in a large number of image formats and can be used as a format converter. Slicer is scriptable and programable. Slicer runs on MacOS, Windows and Linux.

Slicer version 3.6.2 is installed on the Macs in the Lesion Analysis Lab.

Installing Slicer[edit]

The latest stable release is at Mac users also need X11. Most Macs already have X11 installed. If yours doesn't, install XQuartz.

If you don't want to wait 8 to 10 minutes to download the installer from, there are copies of Slicer 4.5 at smb:// Connect to dcmdb and open the installers folder.

To install slicer on a Mac, double click the .dmg file and then copy Slicer to the Applications folder. To install on Windows, run the installer .exe file and run Slicer from the Start menu.

Learning Slicer[edit]

The Slicer training pages contain a number of tutorials and example data sets to get you started. The Slicer Wiki contains a full user's manual. If you require additional assistance, contact Andrew or Nia.

Using Slicer to Convert Formats[edit]

Slicer can load and save data in a wide variety of formats. As well as being able to convert unusual formats, Slicer is also useful when you want to see your data before converting it. There are better options for everyday conversion from DICOM to NifTI.

Format conversions include:

  • DICOM to NiFTi
  • NiFTi to DICOM
  • NiFTi to NRRD
  • NRRD to NiFTi

Use File->Add Volume option to load the DICOM, NRRD or NiFTi data you are converting. When used to load DICOM files, this option includes a DICOM browser. Click for detailed documentation. Slicer does not appear to be able to read in JPG, PNG or TIFF images.

Use File->Save to write the data in the format you are converting to. Click for documentation. Many of these conversions are lossy and will not contain all of the header information present in the original. Not all conversions work. I tested several 'oddball' conversions with version 3.6.3 and they don't appear to work at all.