Thanks very much for the prompt reply, I have one more question mind you, is it possible to set the mapparameters.thresh to a negative number, say -100 (for example) to catch any deactivation as well as activation in the data? When I try it, it seems to stick at threshold of 0?
Cheers
Paddy
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of MCLAREN, Donald
Sent: 17 August 2012 17:21
To: [log in to unmask]
Subject: Re: [SPM] Changing the default threshold in peak_nii
Patrick,
See below.
On Fri, Aug 17, 2012 at 9:55 AM, Patrick Ross <[log in to unmask]> wrote:
> Hi,
>
>
>
> I'm using the peak_nii toolbox and can't work out how to change the
> default corrected threshold of p<0.05?
Set mapparameters.threshc to a new number. This will compute the corrected values for the new corrected p-value. It should be noted that this does not change the thresholding of the data, but rather allows the minimum cluster or minimum peak value to be determined. The thresholding is actually based on the uncorrected value. To threshold with the corrected threshold, you would need to rerun peak_nii with the appropriate thresh or cluster values. For example, say you came up with 100 voxels at p<0.001 to reach cluster FWE. Then you could rerun peak_nii with 0.001 for the thresh and 100 as the cluster.
I can change the T statistic threshold ok,
> but is there a way to say run the program at an uncorrected p<0.001
> threshold?
Yes. mapparameters.thresh=0.001 (in this case you also need to specify the degrees of freedom for the program to compute the t-statistic in the actual image.
The manual says that the mapparameters.thresh input can be the
> T/F statistic, a p-value, or any other number, but entering 0.001 in
> here makes no difference to the default p<0.05 corrected threshold?
As stated above, and to be inserted into the new manual, the corrected threshold only calculates the corrected values. Because of the numerous ways to apply correction (cluster FWE, cluster FDR, voxel FWE, voxel FDR, peak FDR/topological FDR), I have not implemented automatic thresholding. What the program does is compute the p-value or cluster size needed to reach the threshc for each of these methods.
Hope this helps. Let me know if you have any more questions.
>
>
>
> Any help gratefully received!
>
> Thanks,
>
> Paddy
>
>
>
> ----------------------------------------------------------
>
> Paddy Ross
>
> PhD Student
>
> Institute of Neuroscience and Psychology
>
> University of Glasgow
>
>
|