Back

A lasting impact of serotonergic psychedelics on visual processing and behavior

West, C. L.; Bastos, G.; Duran, A.; Nadeem, S.; Ricci, D. A.; Rader, A. M.; Wargo, J. A.; Peterka, D. S.; Van Leeuwen, N.; Hamm, J. P.

2024-07-06 neuroscience
10.1101/2024.07.03.601959 bioRxiv
Show abstract

Serotonergic psychedelics (e.g., psilocybin) have shown potential for treating psychiatric disorders, with therapeutic effects lasting weeks after a single dose. Predictive processing theories posit that psychedelics work by loosening priors or high-level beliefs, including ingrained biases that have become pathological, leading to shifts in bottom-up vs top-down information processing that reconfigure perception, cognition, and mood. Because 5-HT2A receptors, the primary target of psychedelics, are enriched in visual cortices, we investigated whether psychedelics alter visual processing in a manner consistent with predictive processing theories. People who recently (<3 weeks) used 5-HT2A-agonist psychedelics (psilocybin, LSD) exhibited slowed response latencies and increased cortical involvement in generating saccades to targets in predictable locations, along with a generalization of sensory prediction errors (i.e., deviance detection) during passive visual processing. Individuals who recently used a 5-HT1A- selective psychedelic (5-MeO-DMT) displayed similar changes in saccade production, but unaltered deviance detection, suggesting circuit-specific effects. Mice administered DOI (5- HT2A-agonist) exhibited altered deviance detection within primary visual cortex (V1), along with weakened top-down feedback to V1 from higher cortical area ACa. These results concord with the hypothesis that psychedelics shift the balance from top-down to bottom-up in sensory cortical circuits - an effect that persists beyond the acute exposure period.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.