A study published in Nature in 2026 found that psilocybin induces measurable changes in brain connectivity that are not random but follow distinct, context-dependent patterns. Researchers at Monash University conducted experiments with 62 participants who received 19 milligrams of psilocybin and were scanned using MRI while undergoing a series of structured experiences: resting, guided meditation, listening to curated music, and viewing slow-moving clouds. Each participant completed these conditions both under the influence of psilocybin and while sober, with scans taken approximately 80 and 150 minutes post-dosing.

The researchers designed the protocol to avoid cognitive tasks, aiming for ecological validity by capturing the brain’s natural state during an unstructured psychedelic experience. Participants reported the session as among the most meaningful of their lives, with 24 ranking it in their personal top five most significant experiences. Brain activity was analyzed using both traditional methods and a machine-learning approach called CEBRA, which preserved temporal dynamics in neural activity.

Results showed that under psilocybin, connectivity within brain networks weakened while connections between networks strengthened, reducing the brain’s modularity. When eyes were closed, sensory regions lost influence while associative regions gained prominence, suggesting increased dominance of internal mental processes. Notably, the difference in connectivity between eyes-open and eyes-closed states in the visual network shrank by 85 percent under psilocybin, and alpha-band activity—a marker of visual gating—was reduced by nearly half, indicating altered sensory processing.

Using CEBRA, researchers mapped brain activity trajectories across time, identifying four distinct clusters corresponding to each experimental context. A classifier trained on these trajectories could accurately identify which condition a participant was in, and its accuracy correlated with the intensity of participants’ reported subjective experiences. When the activity of key networks—particularly the default mode network and visual network—was swapped between sober and psilocybin states, the classifier’s performance dropped significantly, underscoring their roles in shaping the experience.

Participants who reported greater shifts in self-perception, connection to others, and sense of unity between internal and external worlds showed stronger classifier performance, linking neural patterns to psychological outcomes. The study suggests that psychedelics may temporarily dissolve boundaries between self and environment, not through chaos, but through a structured reorganization of brain networks. While the findings offer a mechanistic framework for therapeutic applications, researchers caution that optimal contexts—such as music or visual stimuli—may vary by individual and require further investigation before clinical deployment.