Forum

Notifications
Clear all

[Solved] Processing parameters for long duration block (+200 seconds)

2 Posts
1 Users
0 Reactions
203 Views
Posts: 11
Topic starter
(@cpickouol)
Eminent Member
Joined: 1 year ago

My interest is in average Hbo and HbR concentrations over task epoch. Data has been processed using block averaging.

I am doing some preliminary analysis on the extracted HRF means for Hbo and HbR on Go/No-Go, and Stop-Signal Tasks. Both measure inhibitory control over a fairly long task epoch (over 200 seconds in my experiement) and are not really event-related as such. It is more of a steady-state analysis. I am concerned that the default bandpass filter 0.01-0.5Hz will flatten out the average concentrations over the course of the task.

Is the a precedent for less conservative parameters? If not, is there justification for less conservative parameters, given that the tasks last well over 1 minute?

If not, are there any suggestions?

Thanks in advance

 

1 Reply
Posts: 11
Topic starter
(@cpickouol)
Eminent Member
Joined: 1 year ago

I have received the following from Gowerlabs re: bandpass parameters and I'm posting here in case anyone else has the same issue

 

"As for the filter, the cutoff frequencies are too conservative. It is probably better to set the range to 0.001 Hz - 0.5 Hz.

The reason is because some of your experiments are quite long and you are interested in the whole time series. From the data you sent me, the longest data set I have is 264 s long (-20s to end of experiment). The longest period of interest would be the length of the block (1/264 = 0.0038 s) and this should not be filtered, so setting the lower cutoff frequency of the band-pass filter to 0.001 Hz should be fine."

Reply
Share:
en_USEnglish