I just computed a design matrix using microtime resolution of 16 and
32. The correlation between the two canonical regressors and
derivative regressors was .9916. This seems to confirm the notion that
increasing the microtime should have little effect. If you decrease
the microtime, it will probably have a larger effect.
Despite the basis functions having different amplitudes at peak, the
actual GLM had roughly the same amplitudes.
On Sun, Apr 26, 2009 at 2:53 PM, Jonathan Peelle <[log in to unmask]> wrote:
> Hi Lili
>
> Just to make sure we're on the same page...The microtime resolution is
> the number of time bins SPM divides each TR into for calculating
> regressor values, and is stored as SPM.T. The microtime onset is the
> bin number, of however many you've specified in SPM.T, from which the
> regressor value is chosen to be entered in the model. The microtime
> onset is SPM.T0. So the default settings of SPM.T=16 and SPM.T0=1
> mean that each TR is divided into 16 time bins, and the value of each
> regressor in the first bin is entered in the model.
>
>
>> The manual recommond not changing the defauft settings (microtime
>> resolution: 16, microtime onset: 1) unless a long TR, while in the Face
>> example (Page 194), these two parameters were set according to the sample
>> parameters (24 for the former and 12 for the later) .
>
> A common approach, especially if slice timing correction is
> implemented, is to set the microtime resolution to the number of
> slices, and the microtime onset to the middle slice. One reason to do
> this is that the middle slice (in time) is never more than TR/2 away
> from any other slice; if you set the microtime onset to the first
> slice, it is nearly twice that (TR) away from the last slice.
> However, I would have thought with a 2-3 second TR, this would not
> make a lot of difference.
>
>
>> I change these settings in my data analysis, using the default settings (16
>> and 1) or use the parameter during data acquiring. This caused some
>> differences in the final result (second level analysis). Then which
>> parameter setting approach is more reasonable?
>
> I would expect changing the microtime parameters to have some effect,
> but probably not a large one. I think that either setting would be
> reasonable. I tend to favor setting the microtime onset to the middle
> slice, and don't know of any reason why this is not a good idea---but
> perhaps there are are some. Changing the microtime resolution I think
> would have less of an effect in this case than the onset, but I don't
> know of any downside to having a larger number (corresponding to more
> time bins) for this.
>
> Hope this helps!
>
> Jonathan
>
--
Best Regards, Donald McLaren
=====================
D.G. McLaren
University of Wisconsin - Madison
Neuroscience Training Program
Office: (608) 265-9672
Lab: (608) 256-1901 ext 12914
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain
PROTECTED HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED
and which is intended only for the use of the individual or entity
named above. If the reader of the e-mail is not the intended recipient
or the employee or agent responsible for delivering it to the intended
recipient, you are hereby notified that you are in possession of
confidential and privileged information. Any unauthorized use,
disclosure, copying or the taking of any action in reliance on the
contents of this information is strictly prohibited and may be
unlawful. If you have received this e-mail unintentionally, please
immediately notify the sender via telephone at (608) 265-9672 or
email.
|