What do dreams tell us about the human brain?

We spend about one-third of our lives asleep and a portion of that time dreaming. But what happens in the brain during a dream? And how can we study something so intimate and fleeting?

An international consortium of scientists has created DREAM, the largest database ever assembled that combines brain activity recordings during sleep (EEG and MEG) with dream reports. It includes over 2,600 awakenings from 505 participants, across 20 different studies. This unique collection allows researchers to study dreams with an unprecedented scale and precision, while addressing one of the biggest challenges in the field: the lack of comparable and accessible data.

Through initial analyses of the database, researchers confirmed that dreams are not exclusive to REM sleep (the phase in which the brain is most active, and the eyes move rapidly). Dreams also occur during NREM sleep, which includes both lighter stages (N1) and deeper ones (N3). Interestingly, when dreams happen during NREM sleep, brain activity resembles wakefulness more than deep sleep, as if the brain was “partially awake.”

In addition, scientists applied artificial intelligence algorithms to analyse brain activity patterns before each awakening. Using these data, they were able to consistently predict whether a person was dreaming at that moment. This innovative approach could, in the future, help identify not only when we are dreaming, but also what kind of experience we are having during sleep.

The DREAM database is freely available at: monash.edu/dream-database, representing a new step in the scientific exploration of human consciousness. This study was published in the journal Nature Communications, in the article A dream EEG and mentation database, as a part of the research project 91/20 - Mentation report analysis across distinct states of consciousness: A linguistic approach, led by Giulio Bernardi and supported by the Bial Foundation.

ABSTRACT

Magneto/electroencephalography (M/EEG) studies of dreaming are an essential paradigm in the investigation of neurocognitive processes of human consciousness during sleep, but they are limited by the number of observations that can be collected per study. Dream research also involves substantial methodological and conceptual variability, which poses problems for the integration of results. To address these issues, here we present the DREAM database – an expanding collection of standardised datasets on human sleep M/EEG combined with dream report data – with an initial release comprising 20 datasets, 505 participants, and 2643 awakenings. Each awakening consists, at minimum, of sleep M/EEG (≥ 20 s, ≥100 Hz, ≥2 electrodes) up to the time of waking and a standardised dream report classification of the subject’s experience during sleep. We observed that reports of conscious experiences can be predicted with objective features extracted from EEG recordings in both Rapid Eye Movement (REM) and non-REM (NREM) sleep. We also provide several examples of analyses, showcasing the database’s high potential in paving the way for new research questions at a scale beyond the capacity of any single research group.