help
help > RE: Negative Connectivity
May 6, 2015 07:05 PM | Alfonso Nieto-Castanon - Boston University
RE: Negative Connectivity
Hi Kaitlin,
If I am understanding correctly, the typical way to help you interpret between-condition differences is the following:
1) run your original seed-to-voxel analyses looking at differences in connectivity between the two timepoints (select 'time1' and 'time2' conditions and enter a contrast [-1 1]). In the results explorer select 'two-sided' (if you want to find both positive- and negative- changes in connectivity) and set the desired thresholds, and then click 'save mask' to create a mask file with your significant clusters (areas where the connectivity with the seed region has changed across time)
2) run another analyses now selecting 'time1' and 'time2' conditions and entering a contrast [1 0;0 1]. In the results explorer click on 'Display Values', then select the option 'clusters defined from a different analysis' and select the mask file that you created in step (1). That will display a barplot with the connectivity values (within the clusters obtained in step (1)) separately for the 'time1' and 'time2' conditions, so you can easily evaluate there whether the found differences relate to increases in positive connectivity strength (e.g. both 'time1' and 'time2' effects are positive and 'time2' is larger than 'time1'), increases in negative connectivity strength (e.g. both 'time1' and 'time2' effects are negative and 'time2' is larger -more negative- than 'time1'), etc.
Hope this helps
Alfonso
Originally posted by Kaitlin Cassady:
If I am understanding correctly, the typical way to help you interpret between-condition differences is the following:
1) run your original seed-to-voxel analyses looking at differences in connectivity between the two timepoints (select 'time1' and 'time2' conditions and enter a contrast [-1 1]). In the results explorer select 'two-sided' (if you want to find both positive- and negative- changes in connectivity) and set the desired thresholds, and then click 'save mask' to create a mask file with your significant clusters (areas where the connectivity with the seed region has changed across time)
2) run another analyses now selecting 'time1' and 'time2' conditions and entering a contrast [1 0;0 1]. In the results explorer click on 'Display Values', then select the option 'clusters defined from a different analysis' and select the mask file that you created in step (1). That will display a barplot with the connectivity values (within the clusters obtained in step (1)) separately for the 'time1' and 'time2' conditions, so you can easily evaluate there whether the found differences relate to increases in positive connectivity strength (e.g. both 'time1' and 'time2' effects are positive and 'time2' is larger than 'time1'), increases in negative connectivity strength (e.g. both 'time1' and 'time2' effects are negative and 'time2' is larger -more negative- than 'time1'), etc.
Hope this helps
Alfonso
Originally posted by Kaitlin Cassady:
Hi Alfonso,
Thanks for the helpful response! I was actually specifically wondering how would I actually evaluate whether the found differences relate to positive-connectivity or anti-correlations between time points. Is there a certain contrast that I should use in the "between-conditions" (second-level) contrast to test for negative connectivity? For instance, if I wanted to examine changes in positive connectivity between time points 1 and 2, I would enter "[1 -1]" which corresponds to brain regions that exhibit decreases in positive connectivity from time point 1 to time point 2. However, how would I look at increases/decreases in negative connectivity between these time points? I hope this makes sense!
Thank you again!
Thanks for the helpful response! I was actually specifically wondering how would I actually evaluate whether the found differences relate to positive-connectivity or anti-correlations between time points. Is there a certain contrast that I should use in the "between-conditions" (second-level) contrast to test for negative connectivity? For instance, if I wanted to examine changes in positive connectivity between time points 1 and 2, I would enter "[1 -1]" which corresponds to brain regions that exhibit decreases in positive connectivity from time point 1 to time point 2. However, how would I look at increases/decreases in negative connectivity between these time points? I hope this makes sense!
Thank you again!
Threaded View
| Title | Author | Date |
|---|---|---|
| Kaitlin Cassady | Mar 24, 2015 | |
| Alfonso Nieto-Castanon | Apr 9, 2015 | |
| Kaitlin Cassady | Apr 15, 2015 | |
| Alfonso Nieto-Castanon | Apr 17, 2015 | |
| Kaitlin Cassady | Apr 23, 2015 | |
| Alfonso Nieto-Castanon | May 6, 2015 | |
| Kaitlin Cassady | May 8, 2015 | |
| Alfonso Nieto-Castanon | May 20, 2015 | |
| Kaitlin Cassady | May 6, 2015 | |
| Kaitlin Cassady | Apr 7, 2015 | |