In 20 subjects of both gender we examined the question whether electro
physiological correlates of mental imagery can be detected using ampli
tude/coherence analysis. 40 concrete words denoting objects were prese
nted to the subjects both visually and acoustically for 2 s each, with
the only instruction of attentively perceiving the stimulus. The care
ful interview afterwards revealed that in some subjects hearing or rea
ding a word had provoked prompt, automatic and detailed mental images
of the object, whereas in others not. This differentiation (visualizer
s - non visualizers) was used to make a comparison between EEG changes
elicited by visualization of a word and EEG changes elicited by perce
iving the word without visualizing it. Since a subject with visual ima
gery while reading the words not necessarily had visual imagery while
hearing them, both groups are independent in this respect; thus differ
ences in the EEG changes of visualizers versus non-visualizers for one
task (visually) need not necessarily be the same in the other (acoust
ically). If we nevertheless would find repeating differences, this wou
ld strengthen the evidence that they reflect the process of visualizat
ion. The EEG was recorded according to the 10/20 system with respect t
o averaged signals of both ear lobes. Amplitude and coherence were com
puted for the frequency ranges theta to beta 3; significant de- and in
creases between each task (visually, acoustically) and a resting condi
tion with eyes opened were mapped. Comparing the maps ''acoustically -
eyes opened'' and ''visually - eyes opened'' reveals different EEG ch
anges for the two modalities; dividing each task into two subgroups of
''visualizers'' and ''nonvisualizers'' reveals differences between th
e groups for each of the tasks. We here only report about the differen
ces between these two subgroups: 1.) During reading as well as during
listening to words the group of subjects in which the words had elicit
ed mental images in the mean shows higher frontal coherence increases
than the control group (fig. 2). 2.) A statistical group comparison re
veals significantly higher beta-amplitude in the occipital region in t
he group of visualizers; this equally applies for both tasks (visually
, acoustically) (fig.4. 3). beta 1-coherence in the group of visualize
rs is significantly lower in the occipital region in both tasks (fig.
3). These results show that mental imagery is correlated with frontal
as well as with occipital EEG changes and therefore support other find
ings, demonstrating that mental visualization is partly evolved in vis
ual areas and partly in frontal, association areas.