>Date: Thu, 30 Nov 2000 11:45:23 -0500
>To: "Joseph T. Devlin" <[log in to unmask]>
>From: Badreddine Bencherif <[log in to unmask]>
>Subject: mask and search volume and resels
>Cc:
>Bcc:
>X-Attachments:
>
>Dear Joseph;
>Thank you for your help. I applied the mask (obtained from n=42) as
>you suggested on data (n=15) to get the same search volume because
>as you pointed out number of zeros were different between both data
>(giving then different search volume).
>Didine
>
>
>>Dear Badreddine,
>>
>>> In a large group of n=42 I have
>>> search volume = S= 536352
>>> 205.5 resels= R
>>
>>> a subset of 15 subjects (of the 42)
>>> S= 651480
>>> 294.6 resels
>>
>>> 1) any explanation for the large difference in S. Data are identical
>>> and I use the same threshold for both.
>>
>>I'm guessing that the difference may relate to the presence of zero
>>values in your data. SPM generate an implicit mask to ignore any
>>voxels which have a zero value in any of the subjects. So if you
>>include more subjects, you may find a smaller search area as additional
>>voxels have been excluded. Presumably this is an asymptoting function
>>though...
>>
>>The way to check this is to compare the masks generated by your two
>>analyses. If there are more excluded values in the mask from the 42 subjects
>>than the mask from 15 subjects, this is likely to be the problem.
>>
>>In this case it should be easy to make the comparison you're
>>interested in. Go to the 15 subject analysis results and use the
>>small volume correction (SVC) button and chose the file option. Then
>>choose the mask from the 42 subjects. That will explicitly include
>>the same set of voxels as the previous analysis.
>>
>>BTW, this is the only possibility which occurred to me but there may
>>be others. It's definitely worth checking the two masks to see if
>>this applies or not.
>>
>>Best of luck,
>>Joe
--
Badreddine Bencherif, MD
Department of Radiology
Division of Nuclear Medicine
Johns Hopkins University School of Medicine
601 N. Caroline St. / JHOC 4230
Baltimore, MD 21287-0855
Phone: (410) 614-2787
Pager: (410) 283-2050
Fax: (410) 614-1977
email: [log in to unmask]
|