Source
Jakub Vohryzek, Joana Cabral, Peter Vuust, Gustavo Deco and Morten L. Kringelbach (2022), “Understanding brain states across spacetime informed by whole-brain modelling”, Philosophical Transactions of the Royal Society A 380, 20210247, LIT-080, read in full in NOTE-054.
What was actually shown
Nothing new; this is a review. What it assembles:
The description. Time-varying functional connectivity is summarised as a repertoire of substates with fractional occupancy, dwell time and transition probabilities (§2, Fig. 2a). Connectome harmonics, the eigenvectors of the structural connectome’s graph Laplacian, give a basis for activity patterns, and the slowest reproduce resting-state networks.
The models (§3). Whole-brain models fit empirical functional connectivity and its dynamics best with small-world topology, weak coupling and local nodes at the edge of a supercritical Hopf bifurcation with noise. Structured noise on structural connectivity explains static connectivity but not its non-stationarity, which is the stated reason for nonlinear local dynamics.
Psychedelics (§6a, cited). Broader connectome-harmonic repertoire under LSD and psilocybin; an enhanced repertoire of substates under psilocybin; higher MEG signal diversity under LSD, psilocybin and ketamine; within-network disintegration with between-network cohesion.
Depression (§6b, cited). Remitted-MDD patients spend less time in a frontoparietal, default-mode, salience and striatal network and more in a global pattern; two further studies report more within-DMN connectivity and more global synchrony and temporal stability in MDD.
The synthesis (Conclusion, Fig. 6). An n-dimensional landscape of substates; depression and psychedelics placed at opposite ends, deepened against flattened; causal models with receptor maps proposed to find perturbations that move between states.
What could have come out otherwise. Nothing in the review was tested against the alternative. That is why this is Proposed.
What this does not say
- That depression is rigid dynamics. The abstract says it flatly; §6b closes with “further studies will be required”, and one of its three studies is of remitted patients. NOTE-054 advises citing the primary studies, after checking them, not this review.
- That the two states lie on one axis. No landscape is estimated from data, and no study places both.
- That transitions are discrete switches. The substate description is assumed, not tested against a continuous model.
- That the brain is critical. Criticality and metastability are premises taken from cited power-law work, not re-examined against its critiques.
- That whole-brain models can design treatments. The step needs out-of-sample prediction of an intervention’s effect, which is not reported.
Connections
- Rival description: cortical spirals (THEORY-088, from LIT-638). Xu
et al. argue that transitions between functional networks “evolve
continuously across space and time, rather than discrete or sudden
changes between different metastable states”, and LIT-638 declares
rivalsagainst this paper. THEORY-088 takes its spiral-defect organisation as its claim and leaves continuity open, saying a discrete-substate theory, once filed, should be weighed against it and that the two may be compatible descriptions at different grains. So norivalsis declared between the THEORYs; the open choice is QUESTION-079’s. - QUESTION-079 asks whether coordination among an agent’s regimes is discrete switching between metastable states or continuous reorganisation. This is the discrete side, stated as a framing; it answers the question only in the sense of being one of the two descriptions the question weighs.
- LIT-066 (meditation and the sense of self) is a natural neighbour for the REBUS and entropic-brain framing; neither paper cites the other.