help
help > RE: Anticorrelations instead of correlations?
Nov 27, 2016 07:11 PM | Alfonso Nieto-Castanon - Boston University
RE: Anticorrelations instead of correlations?
Hi Michael,
That is curious indeed, could you please provide a few more details about your preprocessing? (particularly version number, any relevant preprocessing and/or denoising options, and whether you preprocessed and denoised inside or outside of CONN). There was an old issue, for example, that could cause something like this but for it to happen you would need to: a) having band-pass filtered -or otherwise removed the average BOLD signal from- your data outside of CONN; b) be using PSC (percent signal change) scaling in Setup.Options; and c) be using a somewhat older version of CONN. If your case fits this scenario then simply changing in Setup.Options the data units to "raw" (instead of "PSC units") and re-running the Setup/Denoising steps should fix this (and of course upgrading to the latest release is also recommended).
Best
Alfonso
Originally posted by michael lifshitz:
That is curious indeed, could you please provide a few more details about your preprocessing? (particularly version number, any relevant preprocessing and/or denoising options, and whether you preprocessed and denoised inside or outside of CONN). There was an old issue, for example, that could cause something like this but for it to happen you would need to: a) having band-pass filtered -or otherwise removed the average BOLD signal from- your data outside of CONN; b) be using PSC (percent signal change) scaling in Setup.Options; and c) be using a somewhat older version of CONN. If your case fits this scenario then simply changing in Setup.Options the data units to "raw" (instead of "PSC units") and re-running the Setup/Denoising steps should fix this (and of course upgrading to the latest release is also recommended).
Best
Alfonso
Originally posted by michael lifshitz:
Hi,
I'm running into a weird problem with a basic seed-to-voxel connectivity analysis. I'm using a seed in the PCC. For most of the runs and subjects I see the expected effect: the PCC seed activity correlates to activity in DMN regions. What's strange is that, for some subjects on some runs, I am seeing that the PCC seed is ANTIcorrelated to all of the regions of the DMN that I would expect it to be correlated with. I've attached an image showing an example of this strange effect from one run. Is it possible I am reading the color bar wrong? Any ideas what's going on here?
Thanks so much!
Michael
I'm running into a weird problem with a basic seed-to-voxel connectivity analysis. I'm using a seed in the PCC. For most of the runs and subjects I see the expected effect: the PCC seed activity correlates to activity in DMN regions. What's strange is that, for some subjects on some runs, I am seeing that the PCC seed is ANTIcorrelated to all of the regions of the DMN that I would expect it to be correlated with. I've attached an image showing an example of this strange effect from one run. Is it possible I am reading the color bar wrong? Any ideas what's going on here?
Thanks so much!
Michael
Threaded View
| Title | Author | Date |
|---|---|---|
| michael lifshitz | Nov 23, 2016 | |
| Alfonso Nieto-Castanon | Nov 27, 2016 | |
| michael lifshitz | Nov 30, 2016 | |
| Stephen L. | Nov 26, 2016 | |