NIG
Child's Nerv Syst (1986) 2:72-79
© Springer-Verlag 1986
Electroencephalography in minor head injury in children T a k a o E n o m o t o 1,, Y u k i o O n o 1, T a d a o N o s e 1, Y u t a k a M a k i 1 a n d K e i i c h i T s u k a d a 2 1 Department of Neurosurgery, Institute of Clinical Medicine, University of Tsukuba, Sakura, Niihari, Ibaraki, 306 Japan 2 Tsukuba Neurological Examination Center, Ibaraki, Japan
Abstract. E E G and CT scans o f 280 cases of m i n o r h e a d injury in children under 15 years o f age were studied. A b n o r m a l i t y on initial E E G was shown in 42.5%. Those who lost consciousness h a d a higher incidence o f abnormality than those who did not, a n d it was higher between 4 and 13 years o f age. The sleep state has much influence on the finding. The patients should be awake or in a light sleep stage. The most frequent a b n o r m a l i t y was slow waves, seen p r e d o m i n a n t l y in the occipital regions, and which tended to d i s a p p e a r m o r e easily than the paroxysmal ones. The E E G s became or r e m a i n e d n o r m a l in 95%, excluding incompletely followed-up cases. There was no case of post-traumatic epilepsy in our series, but 4 cases of post-traumatic early convulsions, in which the EEGs were variable. CT scan disclosed a b n o r m a l i t y in 6 %. Key words: M i n o r head injury - Children - E E G - CT scan - Convulsions.
Cerebral concussion has long been the subject o f controversy. Since the elaborate work of D e n n y - B r o w n a n d Russell [4] was published in 1941, it has not been fully understood, and mild head injury is often not easy to assess clinically, especially in younger children. Therefore, we have asked ourselves what h a p p e n s to those who suffer mild head injury and what E E G changes occur. To assist us in this study, we reexamined the E E G s and CT scans taken in the early stages after mild head injury.
Materials and methods We studied a total of 280 patients under the age of 15, who sustained a minor head injury and underwent EEG and CT scan within a week after the injury in one of our affiliate hospitals from 1977 through 1984. Those who had had any kind of cerebral disorder before, including another head injury, and those who had a definte lesion such as massive hematoma, or depressed * To whom offprint requests should be addressed
fracture requiring operation were excluded from the study. The EEG and the CT scan were carried out successively. The EEG was carried out as a rule without the use of sleep-inducing drugs, except for young children and infants (to whom 0.7 ml/kg tricloryl was administered). The silver electrodes were placed in the 10/20 method and connected to the recorder (Nihonkoden). Eye opening and closure, photic stimulation, and hyperventilation were performed routinely. EEG was repeated when needed at intervals of 1, 3, and 6 months, and 1 year after injury. CT scan was performed with EEG, using the Hitachi-CTH, scanning parallel to the orbitomeatal line without dye infusion and repeated when needed. The children were divided into two cohorts: type I, no loss consciousness, and type II, those who were unconscious for less than 6 h. A questionnaire was sent to the parents requesting information concerning headache, emotional change, post-traumatic epilepsy, and so on. Follow-up period ranged from 7 months to 7 years. As to the criteria for judgment, marked slow waves for age, slow wave atypical in distribution, any epileptiform discharges, paroxysmal slow discharges, and significant laterality were regarded as abnormal. Less conspicuous slow waves and equivocal sharp waves were judged borderline. In the CT scan any abnormal high or low attenuation and distortion or disappearance of the normal structures were regarded as abnormal. Those with definite lesions requiring operative intervention were excluded from the study.
Results The age distribution is shown in Fig. 1. There were 204 males and 76 females. O f the 280 children studied, 200 (71.4%) did not lose consciousness (type I) while 80 (28.6%) did (type II). O f the causes, vehicle accidents were p r e d o m i n a n t (118 cases), followed by falls (92 cases), inadvertent blows (23 cases), sports injuries (22 cases), and violence (19 cases). The incidence vehicle accidents increased in older children, falls in younger children. Sports injuries were seen exclusively in the teenage group.
Case reports Case 1 A 9-year-old boy, while riding on a bicycle, was hit by a car. He fell, hitting his head, but did not lose consciousness. He was seen
73
on the day of accident. Neurological and CT examinations were normal, but the EEG showed a slight degree of slowing in both occipital areas. This cleared in a month (Fig. 2).
days. The CT scan disclosed a linear, high density along the posterior half of the interhemispheric fissure while the EEG was not as remarkable and was interpreted as a borderline record. However, some spike discharges appeared in right frontal region, which progressed to generalized spike and wave complex bursts. The basic slowness became normal 2 months after the injury. One year and 6 months later, the tracing was normal except for marked build-up on hyperventilation (Fig. 4).
Case 2 This 7-year-old girl fell from jungle gym. She did not lose consciousness, but developed epistaxis. Neurological and CT examinations were negative. She vomited continuously over the following few days. The EEG demonstrated marked abnormality. It was repeated in 1 week and after 2 months, and on the latter occasion it was normal (Fig. 3).
Case 4 This 8-year-old boy, while playing, fell on his back and hit his occiput. He complained of headache and nausea and vomited once. On examination he was normal. Although CT scan the next day was not contributory, the EEG showed spike discharge in the left central and right occipital areas independently. The latter disappeared 1 month later, but the left centrotemporal spikes remained unchanged (Fig. 5).
Case 3 This 7-year-old boy slipped and fell on his back, hitting his occiput. On examination he was found to have a subcutaneous hematoma in his occipital region. Headache persisted for a few
'°r
Case 5 This 7-year-old girl was hit by a car while she was riding a bicycle. She was not rendered unconscious. The physical and neurological examinations were both negative and the CT scan taken the following day was normal. The EEG recorded on the same day showed sharp waves in the central and midtemporal areas on the right side. Two months later it spread to the other side, associated with generalized spike and wave bursts seen only in light sleep stages. The basic slowness had improved by that time. The spikes on the right side disappeared 3 months and 3 weeks later, leaving some on the left, which practically disappeared 1 year and 3 months after the head injury (Fig. 6).
¢m 2 o
d z
lO
Case 6 U
l
2
3
4
5
6
7 8 9 10 years of age
11
12
13
14
This 8-year-old female was involved in an automobile accident. She was unconscious for about 20 min. On examination, she was drowsy but did not show focal signs. Radiographs disclosed a linear fracture in the fight temporal region. The EEG recorded on
15
Fig. 1. Age distribution oftlie subjects
OD
1M after
H.i.
Fpl-A1
Fp2-A2
F3 - A I .~,.j,
F4-A2
C3-
,~
~,~J~,,~j,~,~
~ v - - ~ .
A1
~
,
u
-
v
.
~
.
-
~
-
~
~..~,.~,~j~al>~,~
~
-
- ..,
-
•
.
t34- A2 ~ , A J e ~ - ~ , . . ~ . ~ , . ~ , v ~ ~ o ~- A ~ ~ % . ~ . ~ 4 a ~ / L ~ ~ , ~ - ~
FT-A1 ~ F
8
-
A
-
2
~
~
,
~
.
~
.
~
.
.
.
.
• -
- .
-
.
~,~.,.,,
r, ,W'v-,~,~
- ~,~,..,~,
.
,
T3-A1
T4-A2
E C G
- ~ ' - ' - - ~ ' - - - J ~ " - - " ~ ' ~
~
~
"
@
~
~
1see .lS0/~v _ ~
Fig. 2. EEGs of case 1. Initial tracing shows theta-wave admixture in bilateral occipital areas (OD), but this cannot be seen a month later (1M)
74 1D
2M
l W
a f t e r H;I. F p t - A 1
F p 2 - A 2
-
.
.
.
.
.
.
.
-.
-" --
"
-~
._7
_
.
, ~ w , . j v ,
F 3 - A 1
F4
-- ~
"
-- / ]
__'
"
.
t . . . .
11 q "
1J
_
0 3 - ^1
FT-A~
~
-
~
-
.
-
_
T 4 - A 2 - -
.
~,
_
.
-
, ~ x , d b q v
~
. . . . . . . . .
F8-A2 ~ T 3 - A 1
=
-
-
"-.
.
.
'~
.
~
.
.
-
.
.
.
"
~
J
"
_
.
-
_
.
.
.
_
_
_
.
.
"
-J
150N v ISOO
ECG
Fig. 3. EEGs of case 2. Initial recording shows delta-waves in the occipital regions and some theta anteriorly (1D). The delta-waves subsided a week later (1 W) and the theta-component disappeared 2 months later (2M)
1D
1M 2 W
4M2W
8M2W
1Y6M
HV 1 ; ~ . AftAr H !
Fig. 4. EEGs of case 3. Although the initial tracing was borderline (1D), the second record showed some spikes during sleep stage (1M2 W). Two months later awake basic rhythm returned to normal. Marked laterality was demonstrated during light sleep 4 months later (4M2W), This progressed to generalized spike and wave complex bursts during light sleep (SM2W). A year and 6 months later, the tracing was almost normal except for a big build-up elicited by hyperventilation (1 Y6M)
75 1114o
1D
after Fpl-AI
Fpl-A1
Fp2-A2
Fp2-A2
c~-,,,,1~'~',-~~ C4-A2
H.I.
~
'
'
'
'
'
~
~
~
"
/
P4-A2
"
"
"
F4- A2
"v-~
+~Y~ ~
~
r
+
,
+
,
/
"
"
"
"
J
"
,
~
o+-^I ~ - " + ~ + - - " + - + - - ' ~ ~ O2-A2
+
o=-~ ~
I
T3-AI
+~++'-~++"-'+~+"]~
T4-A2
~
"
•
/
~
.
/
~
-
~
~
.t
l
F
~
'
~
*
"
"~-~-
~
1
T3"A11
T5-A1
T6 - A2 ~ " ~ ' ~ ' ~
T4-A2
"
-
~
_
.
-,
.,.~-~>~
~.,~,-,.,_.,-~,~..~-,~
"
|50~v
E C G
-
_
~
+~ec
Fig. 5. EEGs of case 4. The first recording demonstrates centrotemporal spikes on the left and independent occipital spikes on the right (1D). Only the latter disappeared a month later, suggesting its traumatic origin (llMo)
1D
2M 2W
3M3W
1Y3M after
H,I.
//
/
O1-A1
o2-~ ~ ¢ , ~ ~ T v ~ v \ FT" A 1 ~ , ~ . , . / ' ~ ' ~ ,
!! ECG
\
_
i i
i --~--~--
Fig. 6. EEGs of case 5. The initial recording disclosed sharp waves in centrotemporal areas on the right side, together with generalized slow basic activity (1D). Two mouths later, spikes and generalized spike and wave complex bursts appeared (2M2W). These cleared up 1 year and 3 months after the injury (1 Y3M)
76
Fig. 7a-c. Plain CT scans of case 6. A linear, high density along the falx seen on the scans taken on the day of the accident (a), and on the next day (b), cannot be seen 4 days later (c)
the following day demonstrated marked by reduced activity on the left side, which eventually cleared up 1 year later. The CT scan on the same day disclosed a linear high density along the interhemispheric fissure, indicative of subarachnoid hemorrhage. The high attenuation was no longer seen on the 4th day (Fig. 7). Of 280 cases, 122 (43.6%) had normal EEGs, 39 (13.9%) borderline, and the remaining 119 (42.5%) were judged as abnormal. With regard to the severity of the injury, 50% of type II were abnormal while 39.5% of type I were abnormal. Type II was divided into three subgroups, depending on the length of unconsciousness: transient; less than 10 rain; 10 min or more. However, there was no correlation between duration of unconsciousness and incidence of abnormahty. As for the relation to age, it was noted that EEG abnormality rates were conspicuous between 4 and 13 years of age (Fig. 8). The time gap between injury and the first record did not seem to have an influence upon incidence of abnormality, at least within 5 days of injury. The rate of abnormality varied according to the patient's sleep state: 51% of awake and 54% of awake with light sleep records, respectively, were abnormal, while that of sleep record only were negligibly low. This was statistically significant
(P< 0.001). As for the quality of abnormality, nonparoxysmal changes (93 cases, 78%) exceeded the paroxysmal ones (10 cases, 8%) and t6 cases (14%) had both. As a whole, 92% of abnormal cases showed basic changes and 22% paroxysmal ones (Fig. 9). That means 9.3% of all cases had at least some kind of paroxysmal changes. Among the nonparoxysmal group, localized occipital slow waves were the most common finding, followed by generalized slow waves and slow waves with occipital dominancy. Others included frontal rhythmic theta or focal suppression, etc. Among the 10 cases with the paroxysmal changes, the commonest one was generalized epileptiform discharges. Simple activating techniques such as eye opening and closure, hyperventilation for 3 or 4 min, and photic stimulation were routinely employed. Of the three, hyperventilation was the most effective to facilitate recording the abnormality: 1.5% of normal and 13.9% of the abnormal group showed abnormally marked build-up, which is statistically significant (P < 0.01). Of those who responded to photic stimulation, 35% of normal and 48,9% of the abnormal group responded to 3 Hz low-frequency stimulation. This was not statistically significant. The eye opening usually blocked the occipital slow and intrinsic rhythm, but in some cases
% 100~.
= EEG
O----O CT
.-
0
1 2
3
4
5
6
7
8
9 10 11 12 13 14 15 yearso~d
Fig. 8. Age and rate of EEG and CT abnormalities. Note that the rate of EEG abnormalities is high between 4 and 13 years of age while that of CT scan is relatively constant in all age groups
with marked occipital slowing it failed to abolish the slow component. Of abnormal, borderline, and normal groups, 63.8%, 43.6%, and 19.7%, respectively, were followed up with EEGs at least once. Of the 24 cases followed up in the initially normal group, only one deteriorated but soon returned to normal. Of the 17 cases followe&up in the initially borderline group, one deteriorated and then remained unchanged almost 3 years after injury. Of 76 cases followed up in the initially abnormal group, 38 returned to normal and 24 were judged abnormal. The mean followup period of unchanged cases, which were initially abnormal, was 4.9 months and that of the normalized ones was 8.8 months. The clinical flowchart of normalized cases (Table 1) shows gross trends in the recovery course, which were arbitrarily divided into three groups: those recovering within 3months of the injury, those within 1 year, and those requiring years to recover. Association of paroxysmal discharges was found in 19% of those who recovered within 3months, 42.8% of those who normalized between 3 months and 1 year after injury, and in 90% of the prolonged cases. Of the unchanged abnormal cases, 75% had paroxysmal discharges at least at some stage whereas a total of 45% of the normalized group had paroxysms. These figures were significant.
77 100 Cumulative Rate of % n o rma l i z a t i o n
• Basic abnormality .......
El
To
Paroxysmal change
50
,
;03 it
I ;
-g r s t• •s
l' r t
0
1 2 3 4 5 6 7 8 0 1 0 1 1 1
2 4 6 8 1 0 2 2 4 6 8 1 0 3 2 4 6 8 1 0 4 year
years
years
years
after H.I.
Fig. 9. EEG recovery in abnormal case. Abnormalities seen in initial EEG Table 1. Flowchart of recovery of initially abnormal cases. Different trends of recovery are recognized. The cumulative rate of normalization was calculated, together with incomplete recovery and unchanged initially abnormal cases, incomplete follow-up cases excluded
Type of abnormality
No. of cases
Nonparoxysmal Occipital slow Occipital dominant slow Generalized slow Other
93 (78%) 41 16 23 13
Paroxysmal Slow Epileptiform General Focal
10 (8%) 1 9 7 2
Nonparoxysmal plus paroxysmal
16 (14%)
The cumulative recovery rate of non-normal cases, with incompletely followed-up cases excluded, was 35% in 3 months' time, 61% in a year's time, 85% more than a year after injury, and a total of 95% judged as normal (with initially normal cases included). The CT scan demonstrated abnormalities in a total of 17 cases, of which 15 were compatible with subarachnoid hemorrhage, like the one shown in Fig. 7. One case was suggestive of hemorrhage in the corpus callosum and the remaining one showed brain swelling. An associated EEG abnormality was present in 10 of 17 cases (59%), which is not especially high compared with that of type II head injury.
With regard to the questionnaires, 53.9% were answered. There were 4 cases of early epilepsy, but no true post-traumatic epilepsy was encountered. Eight cases had previous febrile convulsions, of which 6 showed abnormal EEGs, but only 2 cases showed paroxysmal discharges. Prolonged mild headache persisted in 7 cases.
Discussion
Cerebral concussion, transient unconsciousness with full recovery, was a mysterious p h e n o m e n o n until 1941 when Denny-Brown and Russell suggested direct mechanical action to the n e u r o n as the causative factor producing concussion. In the same y e a r Williams and Denny-Brown [26] reported on E E G changes during and after concussion, in which it was said that initial quiescence was followed by slow waves and that the essential change was central paralysis. Slow waves were merely a reflection o f recovery. They stressed the role of acceleration in the production o f concussion and the paucity o f histological changes. In 1944, W a l k e r e t a l . [23], however, showed experimentally that the essential physiological changes o f concussion were initial overexcitement o f the central nervous system, followed by increased frequency and flat waves. These divergent theories p r o v o k e d discussions that favored the transient paralysis theory. Dawson [3] in 1945, studied E E G s in h u m a n head injury. He described m a r k e d slow waves and generalized
78 suppression in the early stages as indicators of poor prognosis. He then touched upon the susceptibility of the childs brain to epileptic discharges. Of his cases, 24% showed epileptic discharges. From the histological point of view Rand and Courville, in 1934 [17], reported end-bulbs seen in nerve fibers after injury. This, however, had not been given much attention until recently. Holbourn [11], in 1943, introduced the concept of shear strain as a mechanism of head injury, which was later adopted by Strich [20, 21]. Groat et al. [9] produced experimental concussion and verified cell loss in the brain stem interneurons and the pyramidal cells of the motor strip. More than 10 years after this report, Strich, in 1956 and 1961, reported her elaborate analysis of injured human brains [20, 21]. In her reports, she emphasized the importance of shearing strain acting on the nerve fibers and, therefore, on the white matter of the brain (quoting Holbourn and Denny-Brown). She extended her theory to mild head injury and discussed the possibility of incomplete disruption of the nerve fiber as being the cause of concussion. Her theory of white matter degeneration was supported by later reports [1, 14, 15], of which Oppenheimer's showed a definite case of minor head injury with diffuse white matter degeneration. Wei [25] focused his interest on the vessel, reporting endothelial changes of the cortical vessels subsequent to mild head injury in experimental animals. In the field of clinical neurophysiology, Aird and Gastaut [2], in 1959, analyzed a huge amount of cases with occipital slow wave, reporting a significant incidence of sinusoidal 3-4 Hz activity after head injury. These EEG changes were more meticulously analyzed by Silverman [19] in 1962, especially in children. The occipital slow waves seen even in negligible injury were considered by him to be an essential change of head injury. Although he did not follow the cases up he suggested that recovery occur in a month's time. This was further studied by Landau-Ferey, who observed recovery in a week's time [12]. The occipital slow wave was quantitatively analyzed and adapted to clinical evaluation by Mizrahi and Kellawayin 1984 [13]. Although the clinical significance of post-traumatic occipital slow waves was established, in experimental work structual changes were found to play a major role in producing the concussive syndrome by Gennarelli [8] and Povlishock [16]. Since functional cholinergic excitation in the pons was proposed by Hayes [10] in 1984 as a mechanism of concussion, the pathophysiology of concussion has once more become complex. As for the present cases, the overall rate of abnormality of the initial record was 42.5%, exceeding that of the CT scan (abnormal in 6% of total case). These figures seem to us reasonable because the CT scan only detects morphological changes. This finding, however, is important and, according to Dolinskas [5], is indicative of subarachnoid hemorrhage. It supports the role of vascular changes in producing concussion and the post-concussion syndrome [25]. On the other hand, EEG detecs functional changes, so
that these positive rates are reasonable, when compared with previous reports [18, 19, 22]. According to Mizrahi [13] the severity of injury and EEG changes should correlate positively. However, ours failed to show this, partially because we studied the mildest form of injury. The age preponderance of EEG abnormality shown in our study was partly attributable to the administration of sleep-inducing drugs, which make judgment difficult. The awake record is mandatory to find the basic abnormality. Light sleep recordings are necessary if one wishes to find paroxysmal changes. The low incidence of abnormalities in the teenage group may be explained by structual maturation of the brain and its coverings. Almost all of the cases judged to be abnormal showed some basic slowing. Therefore, it seems safe to say that the essential change stemming from mild head injury is the occipital dominant basic slowness of various degree from pure delta wave to alpha wave with some delta or theta component, irrespective of the site of injury. Of paroxysmal changes in a total of 280 cases, 9.3% is not significantly high when compared with the normal population [7]. Although it is true that some cases with paroxysms improved as time lapsed, it is difficult to relate the changes (especially the Rolandic ones) to the head injury. As for the activating technique, hyperventilation was effective in eliciting abnormalities, especially in equivocal cases or during convalescence. It is noteworthy that some cases with slow basic rhythm responded only to lowfrequency photic stimuli and not to approximately 9 Hz. This could not be substantiated with concrete figures, however. In most of the cases eye opening blocked the occipital slow waves, suggesting extracortical origin of slow waves [24]. The failure of eye opening to extinguish extreme delta waves may implicate cortical involvement. Only one of our nonabnormal cases deteriorated to abnormal and stayed unchanged. This indicates the high reliability of the initial EEG in predicting the prognosis. Regarding the recovery rate of the abnormal group, it became clear that the high incidence of paroxysmal discharges played a major role in preventing the abnormal patients from normalizing. Of the paroxysmal discharges, focal spike was a more stubborn obstacle than the generalized spike and wave complex burst. This is significant (P < 0.05).The longest EEG follow-up period was 3 years and 11 months. A total of 95% of adequately followed-up cases, namely more than a year, were judged as normal. According to the questionnaire there was no definite post-traumatic epilepsy, but there were four cases with early post-traumatic convulsions. Initial EEGs were abnormal in three and borderline in one. The quality of abnoramlity was variable. Even a spike discharge could not predict post-traumatic epilepsy [13]. At this stage it seems to us that concussion is a very subtle mixture of functional and morphological changes. Those with dominant functional impairment can recover quickly, even in a week's time, but those with dominant morphological changes recover rather slowly, more than a year after injury. We must always also keep in mind those
79 children with m i n o r head injuries who are not abnormal, but are not n o r m a l either.
Acknowledgement. The authors are very grateful to Mr. Yoshio
13.
Nara, EEG technician, for his selfless assistance. 14.
References 1. Adams JH, Mitchell DF, Graham DI, Doyle D (1977) Diffuse brain damage of immediate impact type. Brain 100:489-502 2. Aird RB, Gastaut Y (1959) Occipital and posteriorelectroencephalographic rhythms. EEG Clin Neurophysiol 11:637-656 3. Dawson RE, Webster JE, Gurdjian ES (1951) Serial electroencephalography in acute head injuries. J Neurosurg 8:613630 4. Denny-Brown D, Russell WR (1941) Experimental cerebral concussion. Brain 64:93-165 5. Dollinskas CA, Zimmerman RA, Bilaniuk LT (1978) A sign of subarachnoid bleeding on cranial computed tomograms of pediatric head trauma patients. Radiology 126: 409-411 6. Dow RS, Ulett G, Tunturi A (1945) Electroencephalographic changes following head injuries in dogs. J Neurophysiol 8:161-172 7. Fukushima Y, Kawaguchi S, Ohsawa T, Onuma T (1973) A study of EEG abnormality in normal children. Folia Psychiatr Neurol Jpn 27:106-115 8. Gennarelli TA, Thibault LE, Adams JH, Graham DI, Thompson CJ, Marcincin RP (1982) Diffuse axonal injury and traumatic coma in the primate. Ann Neurol 12:564-574 9. Groat RA, Windle WE, Magoun HW (1945) Functional and structural changes in the monkey's brain during and after concussion. J Neurosurg 2:26-35 10. Hayes RL, Peehura CM, Katayama Y, Povlishock J-T, Giebel ML (1984) Activation of pontine cholinergic sites implicated in unconsciousness following cerevral concussion in the cat. Science 223:301-303 11. Holbourn AHS (1943) The mechanics of head injuries. Lancet II: 438-411 12. Landau-Ferey J, Bour F (1980) Interet pronostique de I'EEG
15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.
26.
dans les jours qui suivent certains traumatismes craniens de l'enfant et de l'adolescent. EEG Clin Neurophysiol 49: 173-180 Mizrahi EM, Kellaway P (1984) Cerebral concussion in children: assessment of injury by electroencephalography. Pediatrics 73:419 425 Nevin NC (1967) Neuropathological changes in the white matter following head injury. J Neuropathol Exp Neurol 26:77-84 Oppenheimer DR (1968) Microscopic lesions in the brain following head injury. J Neurol Neurosurg Psychiatry 31: 299-306 Povlishock JT, Becker DP, Cheng LY, Vaughan GW (1983) Axonal change in minor head injury. J Neuropathol Exp Neuro142:225-242 Rand CW, Courville CB (1934) Histologic changes in the brain in cases of fatal injury to the head. Arch Neurol Psychiatry 31: 527-555 Sekino H, Nakamura N, Yuki K, Satoh J, Kikuchi K, Sanada S (1981) Brain lesions detected by CT scans in cases if minor head injuries. Neurol Med Chir (Tokyo) 21:677-683 Silvennan D (1962) Electroencephalographic study of acute head injury in children. Neurology 12:273-281 Strich S (1956) Diffuse degeneration of the cerebral white matter in sever dementia following head injury. J Neurol Neurosurg Psychiatry 19:163-185 Strich S (1961) Shearing of nerve fibres as a cause of brain damage due to head injury. Lancet II:443-448 Sugiura M, Mori N, Yokosuka R, Yamamoto A, hnamizu H, Sugimori T, Jinbo M, Kitamura K, Kohno H (1981) Head injury in children. Neurol Surg (Tokyo) 9:697-704 Walker AE, Kollros JJ, Case TJ (1944) The physiological basis of concussion. J Neurosurg 1: 103-116 Ward JW, Clark SL (1948) The electroencephalogram in experimental concussion and related conditions. J Neurophysiol 11:59-74 Wei EP, Dietrich WD, Povlishock JT, Navari RM, Kontos HA (1980) Functional, morphological, and metabolic abnormalities of the cerebral microcirculation after concussive brain injury in cats. Circ Res 46:37-47 Williams D, Denny-Brown D (1941) Cerebral electrical changes in experimental concussion. Brain 64:223-238