1. I am currently running a multi-session temporal concatenation melodic that has 272 inputs (92 subjects x 3 sessions - 4 sessions = 272(bad data)). Is this correct?
2. Their epi sequence is an event-related task design with A & B as the 2 conditions. The events are also quite quick. The 3 blocks are randomized, but the events (trials) within each block are not. Is there a way to incorporate the task information?
3. I have 3 groups (schizophrenia, relative, controls) and want to find the differences between them using the group ICA/dual regression approach.
4. Do I need to put anything into the Time Series model/contrasts or Session/subject model/contrasts within the Post-stats section of Melodic? I am currently just running a group ICA without any specification of these models for the data.
5. I plan to take the group spatial maps and enter them as regressors in a dual regression to find the individual spatial maps and time courses. Is this correct?
|