 |
 |
My best guess is that it is something to do with having NaNs in the masked data, and displaying it with trilinear interpolation. What happens if you display it with nearest neighbour interpolation? Best regards, -John > I was able to get spm_mask to work with multiple input images and > thresholds by changing line 81 from > msk=msk + (img>=thresh & finite(img)); > to > msk=msk + (img>=thresh(i) & finite(img)); > > However, there may be something wrong in this too. The result is a mask > that is much more conservative than what I thought would be the equivalent > expression calculated with img_calc > > In image calc I did > output = (i1>100 & i2>0.15 & i3>0.3) .* img_to_be_masked > > Any ideas why this may be?