A new look at brains transformed by psychedelics

Researchers gave 62 psychedelic‑naive volunteers 19 mg of psilocybin and recorded brain activity with fMRI and EEG across four naturalistic contexts - rest, guided meditation, a music playlist, and a clouds video - so participants could have uninterrupted, ecologically valid experiences. Conventional analyses reproduced the familiar picture of increased between‑network communication and noisier signals: within‑network coupling weakened, between‑network coupling strengthened, and overall modularity dropped. Sensory regions lost global influence while associative regions (linked to imagination and sense of self) gained influence. The usual difference between eyes‑open and eyes‑closed brain states largely vanished under psilocybin, with visual‑network distinctions shrinking by ~85% and alpha‑band EEG power (which normally gates visual input) falling by nearly half.

Using CEBRA, a machine‑learning technique that preserves temporal structure rather than averaging across time and people, investigators found orderly, context‑specific trajectories of brain states under psilocybin that clustered by rest, meditation, music, and movie. A classifier could read moments of activity and identify the context, and its accuracy scaled with how meaningful participants rated their session. Swapping sober for drugged activity showed the default mode and visual networks each contributed over 20% of the effect; their reduced differentiation matched reports of self/world “embeddedness.” Next‑day measures showed widespread positive shifts in mood and connection that correlated with classifier signals. The team argues context is a mechanistic variable clinicians can tune, while noting limits: healthy volunteers only, individual variability, and unresolved questions about how to optimize sensory and musical settings.