Neuroimaging & Brain Measures

Neuroimaging and brain measures are not a condition or a treatment; they are the instruments researchers use to see what is happening inside the head. In psychedelic science three families of tool do most of the work. Functional MRI tracks blood-flow changes to map which regions are active and how they are talking to one another. EEG and MEG record the brain’s electrical and magnetic rhythms millisecond by millisecond, which suits the fast, intense states that drugs like DMT produce. And PET uses radioactive tracers to measure specific molecular targets, most importantly how much of the serotonin 2A receptor a psychedelic occupies, or markers thought to reflect synaptic density.

So this page is unlike the condition pages on this site. There is nothing being treated here, and the word "efficacy" does not really apply. What these studies measure is the drug’s footprint on the brain, the acute reorganisation of activity, the molecular targets it hits, and the changes that may linger afterwards. That makes brain imaging both the most mechanistically illuminating and the most easily misread part of the field. It is illuminating because it shows, directly, what a psychedelic does to a human brain. It is misread when a vivid scan in a small group of healthy people is presented as if it were proof that the drug works as a medicine.

The single most important idea to carry through this page is the gap between a brain signal and a clinical benefit. A change on a scan is data about the drug; it is not, by itself, evidence that anyone got better. Whether the brain effects mapped here translate into help for people who are unwell is the question the condition pages, such as depressive disorders, exist to weigh, and most of the cleanest imaging comes from healthy volunteers rather than patients. Imaging tells us what the tool does to the brain; only clinical work can tell us what, if anything, it is good for.

Because there is no condition here, the relevant "standard practice" is methodological: how psychedelic neuroimaging studies are actually run, and what they can and cannot show. The dominant design is a within-subject comparison, the same person scanned with drug and with placebo, often with very few participants but many scans each, because the effects are large enough to detect in small samples and because dense repeated imaging buys statistical reliability that subject numbers cannot. fMRI provides the spatial maps of connectivity and network structure; EEG and MEG capture the rapid dynamics and complexity measures; PET supplies the molecular anchor, receptor occupancy and candidate plasticity markers.

Two practical points follow. First, the typical study is small and mechanistic. Landmark results come from samples of seven to thirty people, frequently healthy volunteers, with a low dose or an active comparator standing in for placebo. That is appropriate for characterising a drug effect and inappropriate for generalising to a population, which is why these findings are best read as descriptions, not conclusions. Second, the methods carry well-known traps, including the temptation to reason backwards from "this region lit up" to "therefore this mental process occurred", and the reproducibility problems that follow from testing thousands of brain features in a handful of people. A dense within-subject "precision imaging" design in seven volunteers[1]Scientific Data (2025), psilocybin precision-imaging drug trial: dense within-subject scanning over two weeks (n=7) is the field’s response to the first problem; honesty about the second is still a work in progress.

The research outlook for psychedelic neuroimaging is unusually busy, because it is doing two jobs at once: deepening the basic science and serving as the proving ground for new compounds. Phase 1 studies increasingly carry brain-measure endpoints from the start, as with the novel 5-HT2A agonist GM-2505, whose dose-related resting-state EEG changes were mapped in healthy volunteers[1]J Psychopharmacol (2025), novel 5-HT2A agonist GM-2505 Phase 1: dose-related resting-state EEG changes (n=48), and the toolkit itself keeps improving, with denser sampling, better complexity measures, and molecular PET tracers for receptors and synaptic markers.

The most valuable direction, though, is imaging being used to discipline the field’s own claims rather than decorate them. The psilocybin-versus-escitalopram dissociation, where clinical improvement and the expected brain signal came apart[2]Am J Psychiatry (2025), psilocybin vs escitalopram: depression improved but emotional-face BOLD response barely changed, is the kind of result that matters most: it warns against assuming that a treatment must work by normalising whatever scan a study happens to measure. Newer work on ketamine takes a different mechanistic route entirely, with AMPA-receptor-density PET that correlates with its antidepressant effect[3]Molecular Psychiatry (2026), [11C]K-2 AMPAR-density PET correlates with ketamine antidepressant effect in TRD, a reminder that "the brain on psychedelics" is not one story but several. The honest outlook is a fast-moving, genuinely exciting science that is slowly learning to separate what it can see from what it can claim.

The neuroimaging landscape is dominated by academic imaging centres and the pharmacology groups attached to them, the kind of units that can run simultaneous PET-MRI or dense within-subject EEG. Much of the foundational work comes from a handful of well-resourced laboratories, and increasingly from defence- and veteran-focused programmes: the most detailed structural and electrophysiological data on ibogaine, for instance, come from a cohort of special-operations veterans with traumatic brain injury[1]iScience (2026), magnesium-ibogaine in veterans with blast TBI: cortical thickening and brain age -1.3 years on T1 MRI (n=22-25), studied with both cortical-oscillation and neural-complexity measures[2]Nature Mental Health (2025), magnesium-ibogaine in veterans with TBI: EEG slowing and reduced neural complexity persisting at one month (n=30). For industry, brain measures are attractive as early, objective signals of "target engagement", evidence that a new molecule is doing something to the brain, well before any efficacy is known.

For an honest broker, neuroimaging is both the field’s most persuasive evidence and its most over-interpreted. A brain scan carries enormous rhetorical weight; a colourful image of a brain "opening up" on psilocybin is far more convincing to most readers than it should be, given the sample sizes and the inferential leaps involved. The responsible posture is to treat this work as genuinely illuminating about what these drugs do to the brain, to lean on it for mechanism and for the discipline it can impose on overclaiming, and to resist the constant pull to read a vivid scan as proof of a cure. Imaging shows the footprint; it does not show the destination.