Print

Print


Hi Charlotte

> I have 6 binary masks of lesions (patients with stroke) and I would like to
> obtain the intersect between them (the part of the mask which is commun to
> all of them).

If you have 6 binary masks and want to find the intersection, you can
just sum them up and find where this summed image is equal to 6 (since
that would indicate the voxels that had a 1 for each patient).

I would:

(a) Select your 6 masks in ImCalc and add them up to created a summed
image (this can also be helpful in displaying the number of patients
who have a mask at each point):

i1+i2+i3+i4+i5+i6


(b) As a second stage, select the output image from above, and see
where it is equal to 6:

i1==6

[If the above doesn't produce the result you want, try i1>5.9;
sometimes rounding errors will produce an empty image when there
should be values that equal 6.]

The output of (b) should be a binary mask of the intersection.

Hope this helps,
Jonathan