help
help > RE: ROI not defined problem
Jul 23, 2012 09:07 PM | Alfonso Nieto-Castanon - Boston University
RE: ROI not defined problem
Dear Nick,
I believe this was one of the bugs that were fixed in the latest version of the toolbox (currently 13k), if you download the latest version this should work fine. Please let me know otherwise. If you have normalized anatomical volumes, simply enter those as your structural images and leave empty the gray/white/CSF ROIs and the toolbox will perform segmentation on the structural images and fill-in these fields. Alternatively you can also have the conn toolbox perform normalization and segmentation of your anatomical volumes automatically as part of the spatial preprocessing steps (when starting a new project respond 'Yes' to the question regarding 'spatially preproces your data' and select the desired preprocessing steps; or see batch.new options in the manual if using batch scripts)
And regarding your question about the additional sound stimuli, I would recommend using the approach in Fair et al. 2007 (A method for using blocked and event-related fMRI data to study“resting state” functional connectivity. NeuroImage 35:396-405) to remove these additional effects from the timeseries before computing connectivity measures. Using the toolbox you can easily do so by including one additional condition (in the 'Setup.Conditions' tab simply add a second condition -in addition to the standard 'All' condition- specifying the onset/duration of the sound stimuli), and then entering this condition in the 'confounds' list (in the 'preprocessing' tab) in order to have the toolbox remove these effects from the BOLD timeseries.
Let me know if this clarifies
Best
Alfonso
Originally posted by Nick Fallon:
I believe this was one of the bugs that were fixed in the latest version of the toolbox (currently 13k), if you download the latest version this should work fine. Please let me know otherwise. If you have normalized anatomical volumes, simply enter those as your structural images and leave empty the gray/white/CSF ROIs and the toolbox will perform segmentation on the structural images and fill-in these fields. Alternatively you can also have the conn toolbox perform normalization and segmentation of your anatomical volumes automatically as part of the spatial preprocessing steps (when starting a new project respond 'Yes' to the question regarding 'spatially preproces your data' and select the desired preprocessing steps; or see batch.new options in the manual if using batch scripts)
And regarding your question about the additional sound stimuli, I would recommend using the approach in Fair et al. 2007 (A method for using blocked and event-related fMRI data to study“resting state” functional connectivity. NeuroImage 35:396-405) to remove these additional effects from the timeseries before computing connectivity measures. Using the toolbox you can easily do so by including one additional condition (in the 'Setup.Conditions' tab simply add a second condition -in addition to the standard 'All' condition- specifying the onset/duration of the sound stimuli), and then entering this condition in the 'confounds' list (in the 'preprocessing' tab) in order to have the toolbox remove these effects from the BOLD timeseries.
Let me know if this clarifies
Best
Alfonso
Originally posted by Nick Fallon:
Dear experts,
I apologise if this is a basic question. I am new to CONN and looking forward to using the software. I am following documentation and I have pre-processed functional images, however, for my structural images I do not have a CSF segmeted image. Therefore I thought I should put in the pre-processed functional images as well as raw anatomical image and leave the Grey matter white matter and CSF ROI's empty (so that segmentation from the raw anatomicals can be done automatically). Unfortunately, this gives an error for each subjects stating that ROI's have not been defined. How can I get CONN to perform the segmentation?
On an unrelated matter, I have resting data with an occasional sound stimuli (to keep subjects away) should I define each sound as a condition or, in your opinions, am I ok to use the entire block to look at resting connectivity.
Thanks in advance for any help or advice, it will be much appreciated.
Best,
Nick
I apologise if this is a basic question. I am new to CONN and looking forward to using the software. I am following documentation and I have pre-processed functional images, however, for my structural images I do not have a CSF segmeted image. Therefore I thought I should put in the pre-processed functional images as well as raw anatomical image and leave the Grey matter white matter and CSF ROI's empty (so that segmentation from the raw anatomicals can be done automatically). Unfortunately, this gives an error for each subjects stating that ROI's have not been defined. How can I get CONN to perform the segmentation?
On an unrelated matter, I have resting data with an occasional sound stimuli (to keep subjects away) should I define each sound as a condition or, in your opinions, am I ok to use the entire block to look at resting connectivity.
Thanks in advance for any help or advice, it will be much appreciated.
Best,
Nick
Threaded View
| Title | Author | Date |
|---|---|---|
| Nick Fallon | Jul 23, 2012 | |
| Alfonso Nieto-Castanon | Jul 23, 2012 | |
| Nick Fallon | Jul 26, 2012 | |