EEG Meets Cortical Spheroids in Psychosis Research

A review published in Nature sets out a two-track strategy for investigating neural impairments in psychotic disorders: scalp electroencephalography in patients, and human cortical spheroids grown in the dish. The pairing is the story. One method reads the living brain; the other rebuilds a piece of it under controlled conditions.

The review's subject matter is deliberately narrow. Psychotic disorders — schizophrenia among them — have long resisted mechanistic explanation. Scalp EEG can show when and where cortical rhythm goes awry, but it cannot say which cell type failed or which developmental step went missing. Cortical spheroids, clusters of neurons derived from patient cells, can.

Neither method, alone, closes that gap.

What does EEG contribute?

Electroencephalography offers millisecond temporal resolution at the scale of the whole cortex. In psychotic disorders, EEG studies have repeatedly documented aberrations in oscillatory activity and event-related responses. The review treats scalp EEG as the phenotyping instrument: it defines, in patients, the electrical signatures that any mechanistic model must reproduce.

That is a modest claim. It is also the correct one. EEG here does not diagnose anything. It characterizes.

What do cortical spheroids add?

Cortical spheroids are three-dimensional cultures of human neural tissue, derived in most protocols from induced pluripotent stem cells. Because they can be generated from patient samples, they carry the donor's genetic background. Researchers can watch cortical development proceed — cell migration, synapse formation, network maturation — and measure it with techniques unavailable in a living brain.

Crucially, spheroids generate their own electrical activity. Recordings from these cultures can be compared, rhythm for rhythm, with the scalp signatures observed in patients. The review positions this cross-measurement as the bridge between bedside observation and cellular mechanism.

This remains investigational, laboratory-stage work. No device in this pipeline is FDA-cleared, and no diagnostic test for psychosis has emerged from it.

Animal data, human data, and the difference

The review's implicit argument is that psychiatric neuroscience has leaned on animal models whose translational yield in psychosis has been thin. Human spheroids circumvent the species problem by being human tissue. They do not circumvent the problem of being a dish: no vasculature, no immune system, no decades of life history.

EEG supplies the living-human anchor. Spheroids supply the controlled-human substrate. The review's contribution is to argue that the two belong in the same experimental design rather than in separate literatures.

What does this change for the reading room?

For clinical neurophysiologists, the honest answer is: nothing yet, and possibly a great deal later. No montage changes. No protocol changes. What changes is the interpretive frame — the understanding that the odd oscillatory finding on a scalp recording may eventually find its explanation in a dish of patient-derived cortex, and that EEG phenotypes collected today are the reference data against which such models will be judged.

That is a reason to record carefully. It is not a reason to reinterpret.

The Nature review suggests a field arranging its instruments before the decisive experiment. The decisive experiment, evidently, has not yet run.