Hi Jonathan,
I think the modulation should only be applied to expanded
locations that do not have tissue (or CBF etc) before transformation.
Or we should set the values in those regions to be zeros. But I'm not
sure whether there is a way to locate those regions.
Thanks
Ze
On 10/28/2010 10:07 AM, Jonathan Peelle wrote:
> Hi Ze
>
>> I have a hard time to understand modulation for voxel wise tissue
>> intensity comparisons. Let's say sub A's hippocampus is half size of sub B,
>> but they have the same intensity. So if our focus is to check whether the
>> intensity of hipp is different between A and B, should the modulation be
>> avoided? otherwise, modulation will result in a spurious intensity
>> difference in the whole hippocampus.
> That's right...and presumably why the default for functional analyses
> has always been not to modulate during normalization.
>
> Best regards,
> Jonathan
>
--
--------------------------------------------------
Ze Wang, Ph. D
Research Assistant Professor of Biomedical Engineering,
Department of Psychiatry,
School of Medicine,
University of Pennsylvania,
3900 Chestnut Street,
Philadelphia, PA 19104, USA
Tel: 215-222-3200 ext 123
Fax: 215-386-6770
|