<FONT face="Default Sans Serif,Verdana,Arial,Helvetica,sans-serif" size=2>Hi all,<br><br>I have been acquiring some EEG/fMRI data of subjects performing a task. However, since I'm new to the methodology I didn't realise that our task was badly designed for this type of experiment. The problem is that the stimuli appears regularly at 1 seconds interval. So when trying to remove the MR gradient artefacts using the FASTR plugin, I will also remove the physiological signal.<br><br>The solution I have made is to insert my own slice triggers markers (instead of volume trigger markers), in order to generate a slice artefact template which is not affected to the physiological signal as the volume artefact template. This almost solves my problem, but the slice artefact template will leave a residual - the RF onset (EPI sequence). I have tried to exclude to epochs include this residual and it makes my
ERP images look pretty good. So I know my slice template somehow works.<br><br>I wonder if anyone experienced a similar problem, or has any idea how to solve this problem (to get rid of the RF onset)? Filter methods are unlikely to work since frequencies are in the EEG space. ICA seems to be able to clean up a bit... <br><div><br>Any idea is mostly welcome!<br><br></div><div>Sincerely,<br>Martin Uppman<br>Karolinska University Hospital<br>Department of Medical Physics, C2:63<br>SE-141 86 STOCKHOLM, Sweden<br>Work phone: +46 8 585 85927<br>Cellular: +46 73 0331110<br></div></FONT>