Some of my images have been taken in patients with gadolinium, so some regions are highlighted because of the contrast effect. The problem is that, in many cases, this produces a bad parcelation of the brain. For instance, in the example you will find attached, the insula is smaller than it should be (parcels blue, dark read and yellow must be included in the insula).

Is there any way to avoid this effect or to correct these high intensities without using manually correction?

Thanks in advance,

Maria.