Ivan Zimine

Ivan Zimine

Audience level:
All
Friday
9:50 a.m.–10:35 a.m.

Interactive exploration of multi-dimensional image data sets with ipython notebook.

Description

Magnetic Resonance Imaging is a very rich medical imaging modality. It often produces large muti-dimensional data sets which need to be explored interactively and in slightly customized ways. Python is becoming a powerful alternative to matlab for doing this. In this talk I will show how ipython notebook can be used for interactive exploration of regional perfusion data.

Abstract

Magnetic Resonance Imaging (MRI) is a very information-rich medical imaging modality. It allows detailed visualization of the anatomy as well as of various physiological processes in humans and animals. Physiological (or functional) imaging requires acquisition of large muti-dimensional data sets. Such data sets often need to be explored interactively and some tweaking of processing and visualization pipeline is often required. This is difficult to do with stand-alone desktop applications, and this is why Matlab is still a tool of choice for many researchers. But Python is quickly becoming a very powerful alternative due to maturity of numerical libraries, and especially due to the development of IPython notebook. In this talk I will show how it can be used for exploring non-invasive regional perfusion data.

The talk will cover the following:

  • Brief introduction to MRI and ASL.
  • MR images IO (nibabel)
  • Static image display
  • Interactive display using notebook widgets
  • Image arrays, dimension juggling, array math
  • Colormaps and true color images
  • Advantages of ipynb as scientific computing platform

Platinum Sponsors:


Silver Sponsors:


Partners: