Print

Print


Dear Linda


> other two questions:
> 1) Does the toolbox (especially the function se_ROIcount2 you 
> previously sent me) resample the T images from 3x3x3 at 1x1x1 (voxels` 
> dimensions of structural images)?

It sort of does. More precisely, like inthe functions "Calculate image means within anatomical ROIs" (se_ROIextract) or "Functional response of anatomical areas (summary)" (se_Achart) the resampling works as follows:
The voxel list to analyse is defined in 1x1x1 mm anatomical space and then the low-res functional images are simply sampled at the corresponding voxels using NN interpolation. Thatb is, a 3x3x3 mm voxel may be counted 27 times / have a weighting factor of 27 in the mean calculation. While this may seem a little counter-intuitive, in particular when counting super-threshold voxels, it is the correct way to use to high-res ROI definitions in low-res images.



> 2) If I use the option "ROIs defined by to 10-percentile of the 
> probability maps" as ROI in voxel count and mean signal intensity how 
> can I check the result of the function looking at the actual functional 
> images? In other words may I have an image of the ROIs defined by to 
> 10-percentile of the probability map, and may I overlap the thresholded 
> functional activations?

Yes, there is a "hidden" function. Try starting the Anatomy Toolbox and then type se_visualiseCentre in the MATLAB command prompt. This gives you the extend of the regions used when selecting "ROIs defined by to 10-percentile of the probability maps".

Hope this helps
Simon