We donšt acquire the T1w image in the same session as the fMRI data in the
HCP. Also, we use spline interpolation to reduce the blurring effects of
resampling. We also use BBR and field maps so we get very accurate
registrations to the T1w image.
Peace,
Matt.
On 2/8/16, 3:33 PM, "FSL - FMRIB's Software Library on behalf of Colin
Hawco" <[log in to unmask] on behalf of [log in to unmask]> wrote:
>Concatenation can lead to a lot of problems. For example, it is not
>uncommon for the mean of the time series between runs to be very
>different. So if you plot the time course for concatenated data, it has
>there big jumps, like a block design but worse. This then in turn has an
>impact on additional processing steps like filtering/detrending.
>
>There are ways around this, for example by normalizing the mean of each
>time series to a common value and adding extra regressors to account for
>runs, but they are not really ideal except in a small number of select
>cases (I have concatenated, out of necessity, but it needs to be done
>with care).
>
>You do raise a good point about possible biases, though I don't think it
>will be a huge issue. Each data set is generally transformed into MNI
>space by the end. In the case of FSL this happens, generally, after the
>statistics are calculated. I'm not 10%, but I don't think a larger
>transformation is going to have a lot of impact on the results.
>
>Best,
>
>Colin Hawco, PhD
>Neuranalysis Consulting
>Neuroimaging analysis and consultation
>www.neuranalysis.com
>[log in to unmask]
>
>
>
>-----Original Message-----
>From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On
>Behalf Of Cameron Craddock
>Sent: February-08-16 3:56 PM
>To: [log in to unmask]
>Subject: Re: [FSL] registering fMRI series
>
>Hey Steve,
>
>Thanks for the quick reply and for the correction. But, when comparing
>data within session, do you not need to worry about how you register the
>series to one another? Is it sufficient to just coregister each to the T1
>(or even MNI) or might this bias the result to favor one session over the
>other? An EPI collected closer to the T1 might require less alignment
>than one collected later due to head motion.
>
>Cheers,
>Cameron
|