The science of lucid dreaming - 'neuroimaging' of lucid dreams

Until as recently as October of last year, it was considered to be virtually impossible to measure an individual’s dream content. However, scientists at the Max Planck Institute of Psychiatry in Munich and the Max Planck Institute for Human Cognitive & Brain Sciences in Leipzig; working with colleagues from the Charité Hospital in Berlin, have now succeeded, for the first time, in analysing the activity of the human brain during dreaming, through the 'neuroimaging' of lucid dreams. This was achieved with the help of lucid dreaming subjects – i.e. people who become aware of their dreaming state and are able to alter the content of their dreams. The scientists, led by Dr Michael Czisch, research group leader at the Max Planck Institute for Psychiatry, were able to empirically compare brain activity of lucid dreamers for the first time ever as they consciously entertained the same thoughts while sleeping and while awake. The scientists then measured brain activity during the dreamed motion and found that it matched the one observed during the real executed movement in a state of wakefulness. 

Methods like functional magnetic resonance imaging have enabled scientists to visualise and identify the precise spatial location of brain activity during sleep. However, until now, researchers have not been able to analyse specific brain activity associated with dream content, as measured brain activity can only be traced back to a specific dream if the precise temporal coincidence of the dream content and measurement is known. Whether a person is dreaming is something that could only be reported by the individual himself upon waking. "The main obstacle in studying specific dream content is that spontaneous dream activity cannot be experimentally controlled, as subjects typically cannot perform pre-decided mental actions during sleep. Employing the skill of lucid dreaming can help to overcome these obstacles", stated Czisch.

The scientists employed six lucid dreamer subjects who were asked to become aware of their dream while sleeping in a magnetic resonance scanner and to report this ‘lucid’ state to the scientists by means of eye movements. They were then asked to voluntarily ‘dream’ that they were repeatedly clenching first their right fist and then their left one for ten seconds. This enabled the scientists to measure the entry into REM sleep – a phase in which dreams are perceived particularly intensively – with the help of the subject’s electroencephalogram (EEG) and to detect the beginning of a lucid phase. The brain activity (blood flow to various regions of  the brain) measured from this time onwards corresponded with the arranged ‘dream’ involving the fist clenching – although brain activity during sleep/dreaming was notably weaker than in the comparable waking state (up to half the strength of waking brain activity, or activity confined to smaller sections of the brain). 

Functional Magnetic Resonance Imaging (fMRI) machine

Activity in the motor cortex during the movement of the hands while awake (left) and during a lucid dream movement (right)

A region in the sensorimotor cortex of the brain (responsible for the execution of movements), was actually activated during the dream. This is directly comparable with the brain activity that arises when the hand is moved while the person is awake. Even if the lucid dreamer just imagines the hand movement while awake, the sensorimotor cortex reacts in a similar way. The coincidence of the brain activity measured during dreaming and the conscious action shows that dream content can be measured. Martin Dresler, a researcher at the Max Planck Institute for Psychiatry, stated: "Dreaming is not just looking at a dream movie. Brain regions representing specific body motions are activated...with this combination of sleep EEGs, imaging methods and lucid dreamers, we can measure not only simple movements during sleep but also the activity patterns in the brain during visual dream perceptions'. Dresler also suggested that: "[t]he lucid dreamer gains insight into a very complex state: sleeping, dreaming, but being consciously aware of the dream state. This may inform us about concepts of consciousness." 

The scientists were also able to confirm the data obtained using MR imaging in another subject using a different technology. With the help of near-infrared spectroscopy, they also observed increased activity in a region of the brain that plays an important role in the planning of movements. Czisch concluded:  “our dreams are therefore not a ‘sleep cinema’ in which we merely observe an event passively, but involve activity in the regions of the brain that are relevant to the dream content”.

Thus, the discovery suggests that lucid dreaming in combination with neuroimaging and polysomnography (a more common form of sleep monitoring) may allow the transfer of more sophisticated "brain reading" tasks to the dreaming state, the researchers say. In other words, it might eventually be possible to predict dreamed content by analyzing patterns of brain activity. However, Czisch was cautious about the general applicability of results, stating that because the results come from a single participant doing a simple, predetermined task, it’s not clear how other people would perform on self-generated dreams. It seems highly unlikely that empirical research would be able to gain any real insight into entire dream plots. The study was published in a Cell Press journal - see Dresler, Imaging transitions in consciousness: Neural correlates of lucid dreaming, Current Biology (27 October 2011).

In another contemporary study, scientists at UC Berkeley have been scanning the blood flow in their own brains in an MRI machine while watching Hollywood movies.They fed the data into a computer programme which correlated brain activity with motion picture information. Soon, the computer had built up an enormous 'rule book' of brain activity, which it could use to decode what the brain was actually seeing. 

The Californian scientists took one massive leap further than their European peers. They put 18 million seconds of random YouTube clips into the computer, giving it a virtual 'paint palette' to work from. Then, using some remarkable programming, the computer pieced together videos of its own creation, aiming to replicate what the subjects were seeing from within the MRI machine.