191

Resuscitation, 13 (1986) 191-201
Elsevier Scientific Publishers Ireland Ltd.

REPETITIVE

J.J. ASKENASYa,

VISUAL

IMAGES IN SEVERE

J. GRUSKIEWICZb,

WAR HEAD INJURIES*

J. BRAUNC and P. HACKETTd

aSWDC at Loewenstein Hospital, Tel-Aviv Medical School, Tel-Aviv, bDepartment of
Neurosurgery at Rambam Hospital, Haifa Medical School, CDepartment of Neuroradiology at Rambam Hospital, Haifa Medical School, Haifa and dLaniado Hospital,
Na tan ya (Israel)
(Received August 22nd, 1985)
(Accepted October lOth, 1985)

SUMMARY

Out of 20 young inpatients who suffered missile head injuries, two (10%)
presented repetitive visual images (RVI). The RVI appeared during wakefulness when relaxed with closed eyes and also at sleep stage I, II and REM, at
sleep onset and at sleep end. When the two patients were compared neurologically and psychologically with the other eighteen, the coexistence of
combat stress syndrome, diffuse brain lesions, right non-dominant associative
area damage and homonymous hemianopia characterised the two RVI
patients. RVI do not appear in the absence of the combat stress syndrome
even in the presence of the other 3 factors. The sleep may be secondarily
contaminated by wakefulness RVI. The presence of disturbed REM temporal
distribution and short REM latency indicate the depressive state of these
patients. It is hypothesised that a similar brain activity in wakefulness and
REM sleep explain the wake dreaming or wakefulness RVI of those patients.
Key words: Repetitive visual images - After combat nightmares -- Hypnagogic hallucinations - Traumatic stress - Temporal lobe - Insular area Striate cortex

INTRODUCTION

The appearance of visual images other than the visual image of regarded
reality, or images recalled at will is an interesting chapter of human pathology.
Since 1806 the state of mind in which a person imagines he sees persons or
*Presented at the 9th International Congress of Physical Medicine and Rehabilitation,
Jerusalem, Israel, 13-l&h
May, 1984.
Abbreviation: RVI, repetitive visual images.
0300-9572/86/$03.50
0 1986 Elsevier Scientific Publishers Ireland Ltd.
Printed and Published in Ireland

192

things which have no external existence to the senses at that time, was
defined as ideal insanity (Arnold, 1806). Esquirol defined the difference
between hallucinations and illusions, the former being a “perception
without
object” and the latter “a false perception derived from a perceived object”
(Esquirol,
1838). The term “images” was used in the last two decades
(Jaspers, 1972). During the awake state most frequently
encountered
are
psychotic (Charlton, 1963), epileptogenic
psychomotor
(Penfield, 1955), or
occipital hallucinations
(Schneider et al., 1961). During the sleep state hypnagogic hallucinations
are the most common, owing to focal epilepsy or narcolepsy appearing at sleep onset and end (McDonald, 1971), nightmares during
REM sleep (Fisher et al., 1970) and night terrors during sleep stages III and
IV (Broughton,
1968). The relationship trauma-nightmares
has been extensively described. Images while trying to sleep (Horowitz et al., 1980), nightmares owing to Vietnam combat stress (De Fazio, 1975) and World War II
nightmares (Brill and Beebe, 1955) were observed and described. Specific
characteristics
of posttraumatic
stress nightmares versus long life nightmares
were also recently underlined (Van der Kolk-Bassel et al., 1984).
Visual images which do not exist in reality and appear in wakefulness as
well as during sleep, are less well recognised. Two soldiers who were admitted
to hospital with severe missile head injuries, during the Lebanese war, presented this feature. The follow-up lasted for two and a half years. The RVI
persisted for a period of 5-8 months, decreasing linearly in frequency in the
following 8-10 months. The RVI were equally distributed
and of equal
intensity during wakefulness and sleep.
METHOD

Reconstruction
of the craniocerebral
injury was achieved by means of CT
scanning. At the end of a 3-month period, a clinical neurological examination
was performed
and compared
with the neurological
state on admission.
Neuropsychological
testing consisted of the Psychiatric
Status Schedule,
MMPI and a wide range of tests which measured visuo-spatial perception,
praxognostic
processes, visuomotor
coordination,
verbal and visual memory,
thought processes such as classification,
categorisation,
conceptualisation,
reasoning and learning capabilities. It was carried out three times at different
periods during the following two and a half years. Prolonged electroencephalographic
recordings were carried out during wakefulness. Polysomnography consisting of one night of adaptation and two nights of monitoring
was performed at 3 months and 12 months following injury. The recording
technique consisted of EEG recording from C3 and C4 (international
lo/20
system) referred to right and left mastoid sites Al-A2,
using gold cup electrodes. The EOG recorded potentials from two Beckman electrodes placed
at the outer canthi of the two eyes and referred to a frontal electrode. The
EMG consisted of two channels, one recording the tonic muscle activity of
the chin muscles and two recording the phasic muscle activity of the anterior

193

tibia1 and medial gastrocnemius muscle through Beckman surface electrodes,
placed with an interspace of 150 mm at the motor point of the muscles.
EKG and respiration were also recorded. The resistance for all the electrodes
was less than 3000 ohms. The data were analysed according to internationally
established classification criteria (Rechtschaffer
and Kales, 1968).
Presentation
Case I. GE aged 27, an electrician, previously in good health, sustained a
high velocity missile head injury on June 22nd, 1982. The CT scan showed
foreign bodies behind the left eye, subdural haematomata
with multiple
intracerebral
haemorrhages
and damage of the right insular area (Fig. 1). The
metallic foreign body penetrated the left temporo-orbital
area, took a diagonal backward course to the right and lodged deep in the right temporooccipital area (Fig. 2). Owing to these injuries the patient was unconscious
with decerebrate
rigidity in all four limbs, with bilateral dilated pupils and
Cheyne-Stokes
respiration. Within a few hours he underwent a major neurosurgical operation
with bilateral frontal burr holes and aspiration of the
intracerebral
and subdural haematomata.
In the course of the first week he
slowly and progressively recovered consciousness,
began to feed himself and
to speak. His left hemiparesis, left amaurosis and left lower quadrant hemianopia became evident. Two weeks after injury disturbing recurring images

Fig. 1. CT scan of G.E.: multiple intracerebral
including right junction area.

haemorrhages

and brain tisue

dam age

194

.-,-..__
0

I

2

3

-__-

4

5

6

1

G. E. 27 y.o.13d4.10.82
Fig. 2. Patient G.E.: schematic reconstruction of the craniocerebral injuries according to
the CT scan. The double head signifies bilateral injury. The dotted line shows the trajectory from the R orbit. In the lower part: schematic representation of sleep stages. The
shaded areas represent periods of REM. The figure shows short REM latency and disturbed
temporal distribution of REM.

appeared. Two recurring sets of images related to his combat experience
would appear during sleep and wakefulness (upon closing his eyes). In the
first set he was walking in formation
at night, there was a loud explosion
to the left, then he fell to the ground. In the second set he saw 8 or 9 bodies
lying on the ground to his right, the corpses being in a very damaged state.
These two images would appear continuously
any time he closed his eyes
during the day (dozens of times a day) and much more noticeably
when
trying to fall asleep and at the end of sleep, less frequently,
they would
waken him from dreaming during the night. As his general state slowly
improved, the frequency
with which these two sets of images appeared
during the day declined. Four months after the injury he saw the images
only around and during sleep. The set of images presenting his own injury
began to prevail over the other set. Five months after his injury images
appeared only about once a week. Six months following the injury his wife
gave birth to twins. From this time on his dreams took on a more positive
hue and he pronounced
the names of his children in his sleep. As he put it
“the combat tension is beginning to dissipate, but instead worry about the
health of my new-born twins is taking over”.

195

Six months following injury a third psychological
examination
showed
that in spite of the left amaurosis and left lower quadrant hemianopia there
was effective visual compensation.
He still had difficulty in memorising more
complex visual material. Stress and anxiety decreased. The MMPI profiles
were elevated on the SC, D and Pa scales. There was cognitive impairment,
and insight into the severity of his disability developed only very slowly and
even after 9 months it was by no means complete. Severe depressive elements
were present.
Electroencephalographic
monitoring during wakefulness in a relaxed state
with closed eyes in periods when the patient presented RVI did show an
increase of beta activity.
Three months following injury a mean of two polysomnograms
showed
severe insomnia pattern with a sleep efficiency of 45.‘7%, 15 awakenings of
less than 5 min and 6 awakenings of more than 5 min. Sleep stage I = 40.8%;
stage II = 21.0%; REM sleep = 37.2% and stage III-IV = 0%. The REM
latency was short and a severe disturbance
of REM temporal distribution
was present (Fig. 2). Very similar visual images to those appearing during
wakefulness with closed eyes appeared in stage I, II and REM sleep mostly
in the first two cycles. No provoked awakenings were performed
but the
subject was questioned at every arousal. Twelve months following injury the
mean of other two polysomnograms
showed: a sleep efficiency of 74.5%,
6 awakenings of less than 5 min and 2 awakenings of more than 5 min. Stage
I = 21%, Stage II = 53.4%, Stage III-IV = 5.2% and REM sleep = 20.40%.
No RVI were observed during the two nights of recording. Three provoked
awakenings were carried out.
Second CT scan performed
1 year following injury showed in the inferior
region of the right frontal pole an area of low density (porencephaly),
atrophy
of the superior region of the right temporo-parietal
lobe (insular area) with
dilatation
of temporo-occipital
horn, dilatation
of the parieto-occipital
sulcus and mild dilatation of third ventricle (Fig. 3). At the time of CT scan
the neurological
deficit was well compensated
but memory and cognitive
disturbances were almost unchanged.
Case II. TE aged 20, a soldier doing his compulsory
military service, in
good health, sustained a high velocity missile injury on June 24th, 1982.
This consisted of a traumatic amputation of the left arm at the shoulder with
severe external haemorrhage, contusion of the left thorax with pneumothorax
and lacerations of the lower limbs. Surgical closure of the chest injury and
disarticulation
of the stump of the left arm were carried out on the first day
following injury. He remained conscious for the first 30 h following his
injury, then suddenly lost consciousness
and developed a left hemiplegia.
The CT scan showed right temporo-parieto-occipital
infarction including the
insular area and left temporo-occipital
infarction,
most probably an air
embolism infarction
(Fig. 4). He also developed a septic focus in the lung
which then became generalised. These complications
began to clear after 9
days. When regaining consciousness his major complaint was of general weak-

196

Fig. 3. CT scan of G.E.: 1 year following injury: porencephaly of the R inferior frontal
pole, atrophy of the superior region of the R temporo-parietal lobe with ex-vacua 1dilatation of temporo-occipital horn and mild dilatation of third ventricle.

Fig. 4. CT scan of T.E.:
pital infarction.

right temporo-parieto-occipital

infarction and left tempo] ro-occi-

197

ness and repetitive visual images. There were two sets of RVI. In the first he
would see his left arm flying off with blood oozing out. In the second, he
would see himself lying down among the corpses to take cover. The two
images appeared continuously
any time he closed his eyes during the day,
dozens of times a day, at sleep onset and at end of sleep and very rarely,
awakening him from sleep. With improvement in his condition, the frequency
of these episodes diminished, so that after 6 months he would see the images
only two or three times per day. During the first 2 months, he was put on a
number of different anticonvulsant,
tranquillising,
antidepressant
and hypnotic drugs. High doses of hypnotic drugs had to be given to overcome his
struggle to delay sleep onset because of the disturbing frequency and intensity
of RVI at this time. As a result of the patient’s attempt to postpone sleep
onset, his sleep, in addition, underwent
a circadian shift of delayed phase
type. His anxiety and tension decreased and his daytime RVI decreased
significantly after he had seen himself in a television feature, which showed
him swimming as an invalid soldier 6 months after injury.
The detailed psychological
examination,
carried out for the first time at 3
months following injury revealed cognitive disturbances.
These included
impairments of memory. Insight into the severity of his disability developed
gradually, and severe depressive elements were present. A continuous
improvement
was pointed out by the second and third psychological
testing,
performed at 7 months and 13 months following injury. The MMPI profiles
significantly improved also between the second and third examinations.
Electroencephalographic
monitoring during wakefulness in a relaxed state
with closed eyes at a time when RVI were present did show fast beta activity.

0

1

2

3

4

5

‘KEI. 19 y.o.tjO.lO.82

Fig. 5. Patient T.E.: schematic reconstruction of the craniocerehral injuries according to
the CT scan. The double head signifies bilateral injury. The stippled area represents the
infarcted area in both hemispheres. In the lower part: schematic representation of sleep
stages. The shaded areas represent periods of REM. REM latency is short and temporal
distribution of REM disturbed.

198

Polysomnograms
at 3 months following injury showed fragmented sleep.
The mean values of 2 night polysomnograms
showed 14 awakenings of less
than 5 min and 3 of more than 5 min, sleep efficiency = 66.8%; stage I =
31.8%; stage II = 48%; REM sleep = 20%; stage III + IV = 0%. Adisturbance
of the temporal distribution
of REM sleep and a short REM latency was
present. The RVI were concentrated
in the first and second sleep cycle. The
mean values of two polysomnograms
performed
12 months after injury
showed: 6 awakenings of less than 5 min and 2 of more than 5 min,
sleep efficiency = 80%; stage I = 19.5%; stage II = 58%; stage III-IV = 7%
and REM sleep = 15.5% with a normal latency and cyclic distribution.
A second CT scan performed 1 year following injury showed a generalised
dilatation of the lateral ventricles wih various areas of cerebral atrophy predominantly in the right temporo-parietal
lobe.
DISCUSSION

The common feature presented by the two brain-injured patients was that
of RVI appearing dozens of times in 24 h during wakefulness when relaxing
with closed eyes and during sleep stages I, II and REM as nightmares. Stages
III and IV of sleep were poorly displayed and did not offer a basis for observation. The visual experience was more concentrated
in the first cycles of
sleep and at sleep onset. In the comparison of lifelong nightmares with posttraumatic stress nightmares it was found that the repetitive recall of combat
events was significant in the latter and their more frequent appearance at the
beginning of sleep was likewise significant (Van der Kolk-Bassel, 1984;
Kardiner, 1959; Fisher et al., 1973; Kales et al., 1980; Fisher et al., 1973).
The repetitive
character
of visual combat experiences
has already been
emphasised (Kardiner, 1959). The high frequency of RVI in stage II was also
demonstrated
(Schlossberg and Benjamin, 1978). It was suggested that traumatic nightmares have intermediate
characteristics
between the two most
commonly
described dream phenomena:
REM anxiety dreams and night
terrors. The present two cases indicate that RVI can appear as an independent
entity of traumatic visual experience due to the following arguments. Traumatic nightmares described in posttraumatic
stress disorder are accompanied
by increased body movements and above all they do not appear in wakefulness. In the present cases no body movements and no muscle potentials
recorded electromyographically
were observed, the images continued as long
as the eyes were closed and ceased with their opening. If the patient wanted
to stop images, he had to open his eyes or engage himself in activity. The
images were not of epileptic origin nor a psychomotor
oneiric crisis, neither
an occipital fit, because eye closure was the trigger and the patients were in
clear consciousness.
These two non-epileptic
features were associated with
normal electroencephalographic
recordings during wakefulness in RVI periods.
Neither were the images hallucinations
because they appeared many times
during daytime wakefulness.

199

These two case histories represent selected anecdotes but when the descriptions indicating RVI in NREM sleep are added to the present predominant
awake RVI it is reasonable to postulate the existence of an independent
clinical entity of visual images.
It may be asked why out of 20 severe war head injured patients these two
patients (10%) suffered of wake-sleep RVI. Certain similarities in the localisation of the damage suggests a possible relationship with the pathophysiologic
mechanism of RVI. On a background of a diffuse bilateral lesion, right deep
damage at the function of the frontal, temporal and parietal lobes, associated
with right striate cortex damage, expressed by left homonymous
hemianopia
was found in both patients. The right hemisphere localisation in the two
right-handed patients raises the role of the non-dominant
hemisphere. Visual
hallucinations
due to lesions in the non-dominant
cortex of the visual association area were already described (Lance, 1976). In Penfield’s studies visual
hallucinations
as part of a seizure occurred mainly in the non-dominant
hemisphere
(Penfield and Perot, 1963). Hallucinations
were provoked by
stimulating the deep Sylvian and per&insular region with deep electrodes in
epileptic patients (Penfield and Perot, 1963). Visual hallucinations were also
obtained by stimulation of the posterior dorsal hippocampus,
amygdal and
hippocampal
gyrus (Horowitz and Adams, 1978). The development of visual
hallucinations
following temporal lobectomy implies the importance of deep
structures (Jensen and Larsen, 1979). These studies indicate the major role
of superficial
and deep temporo-parieto-occipital
structures
of the nondominant hemisphere in the generating process of visual imagery. This region
is especially concerned
with mnemonic constellations
that form a cortical
association area where there is input of multisensory perceptions of a higher
order (Carpenter, 1978). When this area is damaged and homonymous
hemianopia owing to a striate cortex lesion is also present a predisposing anatomical basis is constituted for RVI, combat stress being the trigger factor. Moreover, research on combat reaction, combat neurosis or combat stress did not
show wakefulness RVI as a frequent feature (Archibald et al., 1962; Block,
1969; Hayes, 1969; Salmon, 1972; Belensky et al., 1983). The RVI in these
two patients emerged when consciousness was regained from coma. The full
symptomatology
of posttraumatic
stress disorder, such as recurrent recollections, psychic numbing and flashbacks were present. The presence of the
organic basis, may suggest why out of many traumatic stress disturbed patients
only a few display wake-sleep RVI. The linear regression of the RVI during
a period between 6 and 13 months indicates that RVI belongs to the class of
traumatic stress disorder.
RVI seemed to contaminate
sleep during epochs of increased cortical
activity such as stage I and REM sleep. It is known that there is a relationship
between light sleep and hypnagogic tendency
(Fisher et al., 1970). In the
present cases the first two cycles were those more contaminated
by RVI. If
during the day the persisting images provoked a depressive reaction, during
the night insomnia was a very evident consequence and its linkage to depres-

200

sion was supported by two features: short REM latency and disturbed REM
temporal
distribution
(Mendels and Hawkins, 1968; Vogel et al., 1975).
Depression was also revealed by repeated psychological
tests. In the two
cases the awake RVI was prominent at the beginning. As the.,clinical state
improved the maximum intensity of RVI moved toward the sleep period,
as if the sleep undertook
the charge of recovery. The hypothesis by which
one may explain the wakefulness
RVI consist in the presence of awake
dreaming owing to the similarities in brain activity between REM and wakefulness. It is suggested that the wakefulness RVI may be microsleep equivalents but so superficial that at will,,opening of the eyes interrupts them.
CONCLUSION

The clear cut parallelism between the recovery from the stress combat
syndrome and RVI disappearance
indicates this clinical entity as dependent
on stress. The dominant clinical sign of wakefulness RVI give right to consider this disorder as an independent
clinical entity. The presence of diffuse
brain lesions, right non-dominant
associative area damage and homonymous
hemianopia
may play a role in the RVI which are predominantly
found in
wakefulness.
REFERENCES
Archibald,
H.C., Long, D.M., Miller, C. and Tuddenham,
R.D. (1962) Gross stress reaction in combat: 15 years follow-up, Am. J. Psychiatry,
119, 317.
Arnold, T. (1806) Observation
on the nature, kinds, causes and prevention
of insanity.
Edited London.
Belensky, G.L., Jones, F.D., Marlowe, D.H., Holloway, H.C. and Harris, P. (1983) Behavioral responses
in adaptation
of combat.
3rd Internat.
Conf. on Psychol. Stress and
Adjustment
in Time of War & Peace. Tel-Aviv.
Block, H.S. (1969) Army clinical psychiatry
in combat zone, 1967-1968.
Am. J. Psychiatry, 126, 289.
Brill, N. and Beebe, G. (1955) A follow-up study of War Neurosis. VA Medical Monograph. Washington,
DC, US government
Printing Office.
Broughton,
R. (1968) Sleep disorders:
disorders of arousal? Science, 159, 1070-1078.
Carpenter,
M.B. (1978) Core text of neuroanatomy.
Baltimore and London, Williams’&
Wilkins.
Charlton, M.H. (1963) Visual hallucinations.
Psychiatry,
Q. 37, 489-498.
De Fazio, V. (1975) The Vietnam
Era Veteran.
Psychological
Problems
J. Contemp.
Psychother.,
7, 9-l 5.
Esquirol, J.E.D. (1838) Les maladies mentales. Paris. Ed. Bailliere.
Fisher, C., Kahn, E. and Edwards, A. (1973) A psychophysiological
study of nightmares
and night terrors. Arch. Gen. Psychiatry,
28, 252-259.
Fisher, C., Kahn, E., Edwards, A. and Byrne, J. (1973) A psychophysiological
study of
nightmares
and night terrors, I: physiological
aspects of the stage IV night terror. J.
Nerv. Ment. Dis., 157,75-98.
Fisher, C.J., Byrne, J., Edwards, R. and Kahn, E. (1970) A psychophysiological
study of
nightmares.
J. Am. Psychoanal.
Assoc., 18, 747-752.
Hayes, F.W. (1969) Military acromedical
evacuation
and psychiatric
patients during the
Vietnam War. Am. J. Psychiatry,
126, 5.

201
Horowitz,
M.J. and Adams, J.E. (1978) In: Origin and Mechanisms
of Hallucinations,
pp. 13-22. Editor: W. Kemp, N.Y. Plenum Press.
Horowitz,
M., Wilner, N. and Kaltreider,
N. (1980) Signs and symptoms
of posttraumatic
stress disorder. Arch. Gen. Psychiatry,
37, 85-92.
Jaspers,
K. (1972) General Psychopathology.
Editors:
J. Hoenig and M.W. Hamilton.
Manchester
University Press, Manchester.
Jensen, I. and Larsen, S.K. (1979) Psychoses
in drug resistant temporal lobe epilepsy. J.
Neurol. Neurosurg. Psychiatry,
42, 948-954.
Kales, J., Kales, A. and Soldates, C. (1980) Night terrors: clinical characteristics
and personality patterns. Arch. Gen. Psychiatry,
37, 1413-1417.
Kardiner,
A. (1959) Traumatic
neurosis of war. In: American Handbook
of Psychiatry.
Editor: S.N.Y. Arieti, Basic Books.
Lance, J.W. (1976) Visual hallucinations
in blind fields. Brain, 99, 719-734.
McDonald,
C. (1971) A clinical study of hypnagogic
hallucinations.
Br. J. Psychiatry,
118,543-547.
Mendels, J. and Hawkins, D.R. (1968) Sleep and Depression.
Arch. Gen. Psychiatry,
19,
445-451.
Penfield,
W. (1955) The role of the temporal cortex in certain physical phenomena.
J.
Ment. Sci., 101, 451-456.
Penfield,
W. and Perot, P. (1963) The brain’s record of auditory and visual experience.
Brain, 86, 595-596.
Rechtschaffen,
A. and Kales, A. (1968) A manual of standardised
terminology,
techniques
and scoring system for sleep stages of human subjects.
UCLA Los Angeles, Brain
Information
Service.
Salmon, T.W. (1972) The war neuroses and their lessons. N.Y. J. Med., 129, 698.
Schlossberg,
A. and Benjamin,
M. (1978) Sleep patterns
in three acute combat fatigue
cases. J. Clin. Psychiatry,
39, 546-548.
Schneider,
R.C., Crosby, E.C., Bagchi, B.K. and Calhoun, H.D. (1.961) Temporal on occipital lobe hallucinations
triggered from frontal lobe lesions. Neurology,
11, 172-179.
Van der Kolk-Bassel,
Blitz, R., Burr, W., Sherry, S. and Hartmann,
E. (1984) Nightmares
and trauma: A comparison
of nightmares
after combat with lifelong nightmares
in
veterans. Am. J. Psychiatry,
141, 187-190;
Vogel, G.W., Thurmond,
A., Gibbons, P., Sloan, K., Boyd, M. and Walker, M. (1975)
REM sleep reduction:
effects
on depression
syndrome.
Arch. Gen. Psychiatry,
32,
765-777.