 |
 |
Many thanks for your response, Mark! Best, Stephane Mark Woolrich wrote: > Hi Stephane, > >> I know this point has been addressed several times on the list, but after >> reading carefully the previous threads, I still need some >> clarification as >> to how best to choose the high pass filter cut-off in the case of 2 block >> design studies I'm working on. >> >> In the first case, I have a classic ABAB design with the period >> between the >> onset of 2 identical conditions varying between 52 and 60s. So, following >> the manual, I was tempted to set the HPF cut-off at 60, but then recent >> messages point out that the manual is not up to date and that the current >> recommendation is not to go under 100 >> (https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0808&L=FSL&D=0&P=193438 >> <https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0808&L=FSL&D=0&P=193438>). >> Should I just go with the default 100 then? >> > > indeed. > >> In my second study, we have a total of 6 conditions alternated in >> randomized >> order and repeated twice within each run. Each of the 6 conditions is >> presented in the first half of the block, and then a second time in the >> second half of the block. The period of time between the onset of the 2 >> occurrences of the same condition is therefore not fixed nor predictable, >> but it can be as long as 319s. In this case, should I set the cut-off at >> something like 320? > > It can be tricky to judge at what frequencies your signal will have > substantial power at. The only way to be sure is to load the design > matrix into something like matlab and look at the power spectrum for > each explanatory variable (EV) and choose a cut-off based on that. > > The feat gui also provides you with some information on how much the > EVs are being effected by the high pass filter. This is because by > default the image of the design matrix that gets displayed when you > click on View Design shows the EVs after they have had the high pass > filter applied to them. You can compare this to the design matrix > without high pass filtering being applied to them by deselecting the > "Apply temporal filtering" option on EACH of the EVs in the GLM setup gui. > > Cheers, Mark. > > ---- > Dr Mark Woolrich > EPSRC Advanced Research Fellow University Research Lecturer > > Oxford University Centre for Functional MRI of the Brain (FMRIB), > John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK. > > Tel: (+44)1865-222782 Homepage: http://www.fmrib.ox.ac.uk/~woolrich > <http://www.fmrib.ox.ac.uk/%7Ewoolrich> > > > > > On 5 Jan 2009, at 20:02, Stephane Jacobs wrote: > >> Hello and Happy New Year to all! >> >> I was wondering if anyone had any idea/advice/suggestions on this >> topic now >> that the holidays are over... :-) >> >> Best, >> >> Stephane >> >> >> On Mon, 29 Dec 2008 19:55:41 +0000, Stephane Jacobs >> <[log in to unmask] <mailto:[log in to unmask]>> wrote: >> >>> Hello FSL users and experts, >>> >>> I know this point has been addressed several times on the list, but >>> after >>> reading carefully the previous threads, I still need some >>> clarification as >>> to how best to choose the high pass filter cut-off in the case of 2 >>> block >>> design studies I'm working on. >>> >>> In the first case, I have a classic ABAB design with the period >>> between the >>> onset of 2 identical conditions varying between 52 and 60s. So, >>> following >>> the manual, I was tempted to set the HPF cut-off at 60, but then recent >>> messages point out that the manual is not up to date and that the >>> current >>> recommendation is not to go under 100 >>> (https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0808&L=FSL&D=0&P=193438 >>> <https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0808&L=FSL&D=0&P=193438>). >>> Should I just go with the default 100 then? >>> >>> In my second study, we have a total of 6 conditions alternated in >>> randomized >>> order and repeated twice within each run. Each of the 6 conditions is >>> presented in the first half of the block, and then a second time in the >>> second half of the block. The period of time between the onset of the 2 >>> occurrences of the same condition is therefore not fixed nor >>> predictable, >>> but it can be as long as 319s. In this case, should I set the cut-off at >>> something like 320? >>> >>> Thanks a lot for any help or advice! >>> >>> Best, >>> >>> Stephane >> > -- ::::::::::::::::::::::::::::::::::::::::::::::::::: Stephane Jacobs Postdoctoral Research Associate Human Neuroimaging and Transcranial Stimulation Lab Department of Psychology 1227 University of Oregon Eugene, OR 97403 - USA Email: [log in to unmask] Tel: (1) 541-346-4184 :::::::::::::::::::::::::::::::::::::::::::::::::::