Skip to main content
. Author manuscript; available in PMC: 2019 Jul 1.
Published in final edited form as: Med Image Anal. 2018 Apr 4;47:81–94. doi: 10.1016/j.media.2018.03.013

Figure 2.

Figure 2

Illustration of temporal variability and spatial variability of brain region r. For the ith FCN (i.e., matrix), a functional architecture of brain region r consists of the rth row elements except for Fi(r, r), i.e., Fi(r, :) = [Fi(r, 1), Fi(r, 2), …, Fi(r, r − 1), Fi(r, r + 1), …, Fi(r, n)]T. We combine all functional architectures associated with brain region r into a matrix Fr = [F1(r, :), F2(r, :), …, Fv(r, :)]T. Each column of Fr is a functional sequence associated with brain region r, e.g., F(:)(r, q) = [F1(r, q), F2(r, q), …, Fv(r, q)]T, and Fr = [F(:)(r, 1), F(:)(r, 2), …, F(:)(r, r − 1), F(:)(r, r + 1), …, F(:)(r, n)]. Finally, the temporal variability of brain region r is defined based on the correlation of functional architectures (i.e., rows of matrix Fr) associated with brain region r across different time windows, while, the spatial variability of brain region r is defined based on the correlation of functional sequences (i.e., columns of the matrix Fr) associated with the brain region r.