[Mristudio-users] mristudio : strange ROI image (my message has not been accepted on the list)

[Julie McEntee]

HI Josselin,

I have encountered the same issue- it seems that, in one of the steps used
to transform an ROI template into another space, the ROIs do not remain
intact. That is, as your pic shows, some of the pixels within each ROI end
up elsewhere in the brain.

The only solution I can think of (and I haven¹t tried it yet) is to ³clean
up² the ROI, that is, use the ROI tools to select and ³subtract² the pixels
that are outside the ROI.

Perhaps someone else on the list has successfully accomplished creating
³clean² ROIs from a transformed template and can comment/assist.

Susumu- I¹m wondering if you¹ve examined whether the mean FA, other stats
obtained from transformed ROI templates are affected by the displaced
pixels?

Thanks,

Julie


On 7/22/10 4:54 AM, "Josselin" <j.houenou at yahoo.fr> wrote:

> Hello,
> 
> I am sending this email directly to you because the list refused it, probably
> because of the joint file.
> 
> Using Diffeomap, I use a WMPM_type_II_atlas and registered it on my native FA
> image through LDDMM and Air. First, it looked fine
> But when I looked deeper and tried to extract only one ROI (the first one):
> here is the result I got: (joint jpg file). The core of the ROI seems ok but
> it also includes many alien points. Do you have any idea ?
> 
> thanks,
> Josselin Houenou
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mristudio.org/pipermail/mristudio-users/attachments/20100722/17a542f1/attachment.html