For about a decade, the received story about psychedelics has been one of chaos. Brain networks — the stable patterns of communication that handle vision, attention or the sense of self — loosen and start talking to each other all at once; EEG traces get noisier and more complex. "Think of the networks of the brain as highways," says Devon Stoliker, a neuroscientist at Monash University. "Under psychedelics, these highways break down into many different directions."
In a recent study published in Nature, Stoliker's team designed an experiment to test whether those directions are truly chaotic and random. They gathered 62 people who had never taken a psychedelic, gave them 19 milligrams of psilocybin and scanned their brains, then used machine learning to analyse the drugged and sober scans. What emerged was not chaos.
An uninterrupted trip
Stoliker does not think chaos is a satisfactory explanation. "It never really explained why an individual would have a meaningful experience, why they might have insight, why they might experience clarity, and why these might translate into positive psychological changes," he says.
Volunteers went through the same four-part sequence twice, once sober and once on psilocybin: eight minutes lying at rest, a guided meditation, an 11-minute curated music playlist, and six minutes of eyes-open video of clouds moving across the sky. Each participant did all four inside an MRI scanner about 80 minutes after dosing, and again on EEG about 150 minutes after.
There were deliberately no cognitive tasks. "We wanted ecological validity," Stoliker says. "We wanted to know what the brain was like when somebody was having an authentic, uninterrupted psychedelic experience." Handing someone a task mid-trip, he notes, tends to pull them out of the state — a phenomenon researchers sometimes call grounding.
Blurred boundaries
With eyes closed, sensory regions lost sway while associative regions gained. Connections within each network weakened and connections between networks strengthened; the brain's modularity dropped across all four parts of the sequence. In the visual network, the gap between eyes-open and eyes-closed connectivity shrank by 85 percent — a difference that, when sober, is enormous. Alpha-band activity, normally a marker of the brain gating visual input, was reduced by nearly half. Half of the participants ranked the session among the most meaningful experiences of their lives; 24 put it in their personal top five.
Hidden order
Standard brain-imaging analysis averages twice — over time, and over people — and can miss ordered structures that appear briefly and vanish. Instead, the team fed the moment-by-moment activity of 332 brain regions into CEBRA, a machine-learning tool that compresses the data down to its essential structure while preserving the order of events. That produced, for each individual, a trajectory through a three-dimensional space.
Under psilocybin, Stoliker says, activity became "organized across time and more clearly differentiated by context." The trajectories separated into four clusters corresponding to rest, meditation, music and the movie. A classifier could read a moment of brain activity and say which one the person was in — and its performance scaled with how profound the participants described their experience as being.
The team then replaced one network's psilocybin activity with its sober version and checked how badly the classifier degraded. The default mode network and the visual network each accounted for more than 20 percent of the effect. Their loss of differentiation may explain why participants reported the self and the external world becoming less separate — a state Stoliker calls "embeddedness."
What remains open
The day after the session, participants rated changes in connection, peace, acceptance and creativity on a scale from –100 to 100, and the vast majority reported positive shifts. Turning that into therapy is another matter. The volunteers were healthy people, not patients; identical "embeddedness" scores can hide very different subjective experiences; and nobody yet knows how the room, the playlist or the imagery should be tuned. "Still, finding this level of organization hidden underneath that disorder and chaos has some really good explanatory value," Stoliker says.
