Skip to main content
. 2017 Sep 8;16:643–658. doi: 10.1016/j.nicl.2017.09.003

Fig. 7.

Fig. 7

Two GMMs learned to represent the normal distribution of FLAIR intensities around their corresponding voxel. Top: A model produced at a location near the boarder between GM and WM. Middle: Mean FLAIR training image. Bottom: A model produced at a location within the WM. Note that the model produced from the border between WM and GM has two distinct components representing the two tissue types, whereas the model produced from within the WM contains two very similar components.