Structural-functional similarity corresponds to states of consciousness

Structural-functional similarity increases under anesthesia
Psychedelics decrease structural-functional similarity under high network segregation
Bayesian structural-functional similarity

Structural-functional decoupling is higher in predominantly segregated substates than integrated at baseline, and this effect is made more drastic by LSD (Luppi2021). This effect is consistent with Psychedelics reduce the contribution of coarse-grained harmonicsprevious work by Atasoy2017, Atasoy2018 which observed that LSD and psilocybin reduce the contribution of coarse-grained functional harmonics

Implementation

(ref: Luppi2021)

Correlation between each subject’s functional connectivity matrix corresponding to an integrated or segregated dynamic substate (brain integration and segregation) and the matrix of structural connectivity (e.g. obtained from a group-average template from the HCP)

“Each matrix of functional connectivity was thresholded proportionally using the same density as the structural connectivity matrix, to ensure that both matrices would have the same number of non-zero entries.” (Luppi et al., 2021, p. 4)

“Then, the upper triangular portion of each connectivity matrix (structural and functional) was flattened into a vector, and the Spearman correlation coefficient between these two vectors was computed, as our measure of similarity between functional and structural connectivity.” (Luppi et al., 2021, p. 4)

Alternative

Hamming distance between binarized connectivity matrices (Tagliazucchi2016)

  • Proportion of connections that need to be changed before two connectivity patterns become the same
  • Measurement of dissimilarity

You see Bayesian structural-functional similarity


brain networks
brain integration and segregation
brain network integrity
Loss of consciousness suppresses anticorrelations between brain regions


Q&A