Export to SPM: choosing time windows

1

Hello,

I am exporting my data to SPM12 so I can run a multiple comparison ANOVA test. The baseline for the files is from -200-0ms. After that I want to look at activity every 100 ms starting at 400ms (400-500ms, 500-600ms, 600-700ms, etc.)

Should I add an export process for each of these time segments? Or am I able to export the entire file and then set my time segment preferences directly in SPM12?

Thank you!

1 Like
2

Both approaches are possible:

  • Saving multiple time points in a 4D .nii file (using a large value for the "time downsample factor") maybe after applying a low-pass filter)
  • Saving one 3D .nii file for each time segment.

I think I would go for option 2, as it creates smaller files, and by naming correctly the output files it makes it easier to track which file corresponds to what.
The 4D .nii volumes including a time dimension (option 1) would not include correct time values, therefore in SPM you'd need to refer to them with their indices (eg. ",1" after the .nii file name points at the first volume in the sequence). It is easy to make errors converting a latency into a volume index...

1 Like
3

@Francois do you have a resource/tutorial for running an ANOVA test in SPM12? I was able to export the data from Brainstorm to SPM12 with ease but running the test itself is not very straightforward and I'm having trouble finding a good resource online. Should I be loading my .nii files into Scans? I'm also not sure where to load my .dat files. Any suggestions?

Thank you,
Rawan

4

I'm sorry, beyond the SPM12 tutorials I wrote a few years ago, I'm not very familiar with statistics in SPM...
https://neuroimage.usc.edu/brainstorm/ExportSpm8
https://neuroimage.usc.edu/brainstorm/ExportSpm12

You can start by referring to the SPM12 manual, and then ask questions on the SPM mailing list if you have questions.

1 Like