Dear Peter,

 

Thank you for your kind response.

I understand the rationale of your suggestion, which I assume, is to create a binary mask of the same size as the ROI and mask the thresholded T-map with it to obtain the activated voxels. However, 2 things are still puzzling to me and I would really appreciate some more opinions.

 

First and foremost, how do I create the binary mask as suggested in your last mail? This is also the initial question of my original post. My thought would be, in the case of a spherical ROI for example, to include those voxels whose distance from the sphere center is smaller than the radius. This method, which I am not sure of its validity, might introduce the concern of voxel size as discussed in my original post. So according to your expertise, is there any better way or is my approach acceptable?

 

Second, how do I get the thresholded T-map? I suppose the SPM_T***.img generated after model estimation corresponds to the calculated voxel-wise t-values with no thresholding and correction (if needed) performed.

 

Many thanks and best regards,

                    Ce

 

 

 

----- 原始邮件 -----
发件人：Peter Erhard <[log in to unmask]>
收件人：[log in to unmask]
抄送人：[log in to unmask]
主题：Re: [SPM] voxel sizes and ROI
日期：2012年05月22日 16点54分