On 5/17/2021 10:16 AM, Ellen JOOS wrote:

External Email - Use Caution
Dear Freesurfer team,

I am currently preparing my first fMRI study and used OptSeq2 for creating the trial sequence. Since I am not experienced with the study design in MRI (I mainly conducted EEG studies so far), I would like to have your professional opinion on my design choices and use of your program.

The parameters of my study are the following:

I will conduct a multiband fMRI study with an event-related design that will have a TR of 0.5s. I will present 10 different types of trials that will have a length of 2.5s each, but that will be shown with different amounts of repetitions across the experiment. Within the 2.5s trial time, I included a fixed time period of 1.3s to respond.

As indicated in your recommendations, I added time to the overall presentation time for the "NULL events" that corresponds to the time needed for one of my trial types (as if I would have another trial type). You can find the output of OptSeq attached.

My questions are the following:

1) Did I use the program correctly by having a fixed response time window within the trial duration estimation?

Yes

2) Is the use of the NULL events (as can be seen in the attached file) correct and sufficient to introduce varying amounts of time between the individual trials such that the hemodynamic response can be captured appropriately?

This is a tricky question because you end up not having many nulls relative to your task, so they might look a little like odd balls (eg, at one point there are 7 task trials between nulls). This is more of a question for you and whether you think the psychology is affected by the nulls. I would probably double the null time.

3) In the summary of the OptSeq results, I found an efficiency value of 0.009 and a VRFAvg of 6.24. I saw in your recommendations that the VRFAvg should be 20-40. Is the result of my calcuation too bad to have an efficient recording that allows analyzing the data in an event-related manner?

Are you planning to use an FIR or are you going to assume a shape to the HRF? Probably the latter; if so, then those values are not meaningful.

4) In the actual experiment I will present the attached sequence twice. I have now the options to either present two different sequences (e.g. the two most efficient sequences produced by OptSeq) or to repeat the most efficient sequence twice. Do you have a recommendation? And further, might using the same sequence across participants result in a general ordering effect that might bias my results?

I would use two different sequences and then randomize the order across subjects.

I am very much looking forward to your response and already thank you for your advice and help!

Kind regards,

Ellen Joos
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer