Understanding how psilocybin reshapes the brain has been constrained by small samples, single imaging modalities, and short follow-up windows. A newly released open-access neuroimaging dataset addresses all three limitations simultaneously, potentially accelerating the entire field of psychedelic neuroscience by giving researchers a shared, high-quality foundation to test competing hypotheses.
The PsiConnect dataset captures 62 participants before and after a standardized 19 mg psilocybin dose using an unusually rich methodological stack: multi-echo fMRI, structural MRI, diffusion-weighted MRI, and simultaneous EEG. Rather than relying solely on resting-state scans — the default in most psychedelic neuroimaging — the protocol introduced three ecologically grounded conditions: guided meditation, music listening, and movie watching. Crucially, half the cohort completed an 8-week structured meditation program prior to dosing, creating a built-in factorial design to disentangle psilocybin effects from contemplative practice effects. Behavioral and self-report assessments extend to one year post-administration, providing longitudinal resolution rarely seen outside clinical trials.
This dataset matters less as a single study and more as scientific infrastructure. Most published psilocybin neuroimaging work draws on cohorts under 30 participants with single-modality imaging and follow-ups of days to weeks. PsiConnect's multi-echo acquisition meaningfully improves BOLD signal quality by separating neural signal from physiological noise — a technical advance that should reduce false positives in connectivity analyses. The meditation arm is particularly valuable: the field has long speculated that set and setting modulate psilocybin's neural signature, but controlled designs to test that have been scarce. Limitations to note: the 19 mg dose is moderate rather than high, the sample is likely skewed toward psychedelically and meditatively experienced adults, and as an observational dataset it cannot establish causality on its own. Still, releasing curated, BIDS-compatible data aligned with open-science standards is an incremental but genuinely enabling contribution — one that could compress years of redundant data collection across independent labs.