The dreams that an individual is able to recollect can be synthesized into a visual using computer graphics. The EEG recorded overnight can then be compared to an EEG that is being recorded when the visual is re shown to the individual.
I am not sure what is going to be possible in the future, but even if dreams can be visualized EEG is not going to be appropriate tool for such a purpose. EEG is a time and space averaged electrical signal of the neuronal activity in the cortex with a fairly good temporal but a very bad spatial resolution, therefore important spatial detailes are lost (remain hidden) for the electrodes. No way.
The spatial yield can be improved by augmenting a fMRI with EEG. There is atleast one study that has analyzed brain activity to look for neural correlates in people with lucid dreaming and non lucid dreaming with help of fMRI and EEG.
Neural correlates of dream lucidity obtained from contrasting lucid versus non-lucid REM sleep: a combined EEG/fMRI case study.
Dresler M, Wehrle R, Spoormaker VI, Koch SP, Holsboer F, Steiger A, Obrig H, Sämann PG, Czisch M.
Author information
Abstract
STUDY OBJECTIVES:
To investigate the neural correlates of lucid dreaming.
DESIGN:
Parallel EEG/fMRI recordings of night sleep.
SETTING:
Sleep laboratory and fMRI facilities.
PARTICIPANTS:
Four experienced lucid dreamers.
INTERVENTIONS:
N/A.
MEASUREMENTS AND RESULTS:
Out of 4 participants, one subject had 2 episodes of verified lucid REM sleep of sufficient length to be analyzed by fMRI. During lucid dreaming the bilateral precuneus, cuneus, parietal lobules, and prefrontal and occipito-temporal cortices activated strongly as compared with non-lucid REM sleep.
CONCLUSIONS:
In line with recent EEG data, lucid dreaming was associated with a reactivation of areas which are normally deactivated during REM sleep. This pattern of activity can explain the recovery of reflective cognitive capabilities that are the hallmark of lucid dreaming.
Another study from Science I think. Getting a lot of press.http://www.sciencemag.org/content/340/6132/639.abstract?keytype=ref&siteid=sci&ijkey=PoX6VQkWO8MDg
Recording electrophysiological correlates of dreams is very different from recording dreams. At the present time we are limited to record verbal reports of dream without any notion of when this dream has occur during the sleep. We don't even know if we dream in the same way during all stage of sleep.
Lucid dream is a very particular (and controversial) altered state of consciousness and I am not sure it is a good model to study classical dreams.
Yes, I agree with Ritwick. The paper that he mentioned got lots of attention. As Alain pointed out there are two different things. One is recording the dreams (not just the REM sleep, but waking up the subject and asking questions); another one is electrophysiological decoding of dreams. I have personally done the former but not the later.
Feasibility of EEG based image reconstruction has been demonstrated by Nemrodov and Colleagues of University of Toronto in their published work "The Neural Dynamics of Facial Identity Processing: Insights from EEG based Pattern Analysis and Image Reconstruction". Available at: eNeuro 2018; 10.1523/ENEURO.0358-17.2018