Understanding body planes through MRIs and ItkSnap

The objectives of this homework are:

  1. Get familiar with the software ItkSnap and 3D Slicer.
  2. Understand the three body planes (Transverse, Sagittal, and Coronal)
  3. Introduce the student to the Itk library and the basic IO functionality.
  4. Continue practicing with Jupyter Widgets.

For this homework, you need to make your report in a Jupyter Notebook. If you are not familiar with JupyterNotebooks I encourage you to take this tutorial here.

Body planes and ItsSnap/3D Slicer (10 Pts)

Using this T1-weighted MRI brain image to identify the following regions in the brain:

  • From a Sagittal view:
    • Cerebellum
    • Medulla Oblongata
    • Corpus Callosum
  • From a Coronal view:
    • Optic nerve
    • Sylvian fissure
  • From a Transverse (Trans) view:
    • Lateral ventricle
    • Caudate nucleus

Bellow is a Sagittal schematic view of the brain, you need to Google the other regions of the brain to find out more information:

Brain regions

For each case make a screenshot of the identified region using the proper plane view and mark the region with an image-editing software like Photoshop, Gimp, Paint, or even Powerpoint. If you are not familiar with any of these tools, I suggest you to use Gimp, it is free and open-source. You can find a tutorial here.

This is an example image showing the Pons area from a Sagittal view:

Pons

Itk and IpWidgets (10 Pts)

Install the Itk library with one of the follwing commands in your python environment:

pip install simpleitk
conda install simpleitk::simpleitk

Using the Prostate.mha file inside the TestData folder as an example, create an IpWidget that displays a view of the 3 planes (similar to ItkSnap) and allows the user to select the location of each plane with an IntSlider.

Similar to the image bellow (but I'm sure you can improve all the labels and the layout):

widget

Extra Itk and IpWidgets (5 Pts)

In a new cell of your report, upgrade your widget to display the three planes but this time display each plane with the CORRECT proportions, taking into consideration the pixel sizes. Similar to what you see in ItkSnap.