brain white matter tissue probability map as mask - behaviour of pyradiomics

greetings to community,

i have a doubt:
when i use spm12 to produce a brain white matter segmentation from a T1 MRI scan the image produced by spm has various intensities from 0 to 1 especially at the edges between the white and grey matter as it is a tissue probability. So when using this image as mask, how does pyradiomics behave with regards to which value of intensity to use for cut-off between the classes?
Ideally the mask should be either 0 or 1 at each point.