Pynax stands for Python-n-axis. This is a visualization tool written in Python that provides interactive data exploring GUI.
Pynax 3D viewer
I developped Pynax during my first month as a PhD. In fact, I have made an extensive use of nisl in my projects, to manipulate neuroimaging data. Unfortunately, visualization tools are not yet provided and I had a strong need of interactive plots. Great visualization tools for neuroimaging data already exists. FSLview, MRIcron and Anatomist are great tools but using them requires a dump of my data. Plus, if you want to overlay several activations maps, for example, doing it each time by hand is time-consuming. Nipy provides great vizualisation scripts but they are not interactive and their formatting is rather limited.
Pynax is here to solve these problems. It provides:
- Full matplotlib integration: you can reuse your figure generation scripts
- Interactive layout: you can browse your 3D or 4D data easily
- Great extensibility: you can build in 10 minutes the interactive view that best fits your needs.
However, this is mainly a debug tool. The available views fit my needs and, as I am not a matplotlib guru, they are ugly.
Coloring a parcellation is not an easy task. In fact, each parcel color should be distinguishable from its neighbors'. This problem is known as the four color theorem in a 2D plane, and the k-coloring problem in a _n_dimensional problem.
An easy solution is to color the parcels by selecting random colors in the spectrum and assigning them to the parcels. However, several trials are needed to end up with a suitable result.
Thanks to Pynax, I made up an interface to change the parcel colors by clicking on it. This tool has been really helpful to end up with beautiful figures.