How to control displayed orientation in nilearn plots

Is there a way to control the display convention in nilearn plots (radiological vs. neurological)? I am new to nilearn, and I haven’t found the right plot_stat_map arguments to control this part of plotting yet. Thanks!

AFAIK. I don’t think there is a way to display in radiological convention.

OK, glad I’m not just missing something obvious. Thanks!

Reviving this older thread.

Relatively new to nilearn and considering adopting this tool for displaying brain plots in a more reproducible way. The default plotting view is in neurological convention (left-is-left), and like the users above, I also do not see a way using plotting function parameters to show the images in radiological view (left-is-right). As such, I’ve decided to open an issue:

Open to suggestions/workarounds either within the nilearn framework or perhaps using other approaches. Thanks!

Hi @jclauneuro, not sure if this helps. But as a workaround, could you use nilearn.image.swap_img_hemispheres to just swap the hemispheres?

Thanks @miykael for the suggestion. nilearn.image.swap_img_hemispheres is a workaround but of course needs to be used with extreme caution and with the plotting function parameter annotate = False otherwise the L/R sides are labeled incorrectly. Created a simple notebook example to illustrate this.

In the meantime, @GaelVaroquaux has responded on GitHub and stated that implementing a radiological = False parameter to plotting functions would be worth pursuing.

Yes, you’re right. I also worried a bit about the annotation issue but thought that you could just turn it off. But I see from your example notebook that this is not a good option, especially if somebody forgets to turn it off.

Have fun with implementing this to nilearn if you chose to do it yourself. Everybody there is very nice and friendly, and the whole process should be very smooth.

1 Like