Childs Nerv Syst (2010) 26:755–762
DOI 10.1007/s00381-009-1047-8
ORIGINAL PAPER
Natural history of traumatic meningeal bleeding in infants:
semiquantitative analysis of serial CT scans
in corroborated cases
Matthieu Vinchon & Marie Desurmont &
Gustavo Soto-Ares & Sabine De Foort-Dhellemmes
Received: 30 October 2009 Published online: 28 November 2009
# Springer-Verlag 2009
Abstract
Background The natural history of posttraumatic meningeal
bleeding in infants is poorly documented, and the differ-
ences between inflicted head injury (IHI) and accidental
trauma (AT) are debated. Autopsy findings have suggested
that anoxia also plays a role in bleeding; however, these
findings may not reflect what occurs in live trauma patients.
Purpose We studied the natural history of traumatic
meningeal bleeding in infants using serial computed
tomography (CT) scans in corroborated IHI and AT.
Materials and methods From our prospective series, we
selected corroborated cases (confessed IHI or AT having
occurred in public), who underwent at least three CT scans
in the acute phase. We performed a semiquantitative
analysis of meningeal bleeding using a four-tier scale
(absent, faint, frank, and thick) derived from the Fisher
grading for aneurysmal bleeding in four regions of interest
(convexity, falx cerebri, sagittal sinus, and tentorium
cerebelli).
Results We studied 20 cases: ten IHI and ten AT. Bleeding
was maximal at the convexity initially, then increased along
M. Vinchon (*)
Department of Pediatric Neurosurgery, Lille University Hospital,
59 037 Lille Cedex, France
e-mail: m-vinchon@chru-lille.fr
M. Desurmont
Department of Legal Medicine, Lille University Hospital,
Lille, France
G. Soto-Ares
Department of Neuroradiology, Lille University Hospital,
Lille, France
S. De Foort-Dhellemmes
Department of Neuroophthalmology, Lille University Hospital,
Lille, France
the falx and sagittal sinus, and then along the tentorium.
Decrease and disappearance of blood was variable accord-
ing to the site and the initial quantity of blood. We found no
difference between IHI and AT.
Conclusion Our findings suggest that the primary site of
meningeal bleeding in infantile head trauma is the convex-
ity of the brain; blood cells then migrate toward the midline
following the flow of cerebrospinal fluid circulation and
inferiorly following gravity. The pattern of bleeding in
traumatic cases appears similar in IHI and AT but different
from anoxic lesions.
Keywords Traumatic head injury . Infant .
Shaken baby syndrome . Accidental head trauma .
CT scanner . Subdural hematoma . Pathophysiology
Introduction
Subdural hematomas (SDH) are common traumatic lesions
in infants not only in inflicted head injury (IHI) but also in
accidental trauma (AT). Computed tomography (CT)
scanner is the most widely used imaging modality for the
initial diagnosis and subsequent follow-up of meningeal
bleeding in infants [1, 7]. However, the natural history of
meningeal bleeding in infants is poorly documented
because cases with undisputable trauma, precise dating,
and serial imaging are uncommon. Cases of confessed IHI
are difficult to collect, and cases of corroborated AT
(having occurred in public) are rare.
As a result, the radiological description of meningeal
bleeding in infants has always been mostly qualitative, and
differential diagnosis between IHI and AT has been based
on such features as thickening of the falx [16, 31],
subtemporal hematoma [2, 26], and mixed-density hemato-
756
mas suggesting repeated trauma [15, 16]. These depictions
have often been arbitrary and tainted by circular reasoning
(description of findings in categories based on the same
findings). Other studies have found that features often
considered as indicative of IHI are in fact common in
infants, whatever the cause of trauma [33, 35, 36]. The
temporal evolution of meningeal blood is also poorly
documented; however, knowledge of time-related altera-
tions might be useful in forensic investigations for the
dating of trauma and the diagnosis of repeated assault. In a
previous study, we found that dating of trauma could be
based on the pattern of blood in CT and magnetic resonance
imaging (MRI) scans [34].
Another question is the origin and pathophysiology
of meningeal bleeding. The classical hypothesis is
tearing of the corticodural bridging veins caused by
shaking [14, 21–23], but this theory has been challenged
by recent advances in anatomy [19] and pathology [12, 27,
28]. Several authors now consider that the bleeding
originates from the dural venous plexus close to the
sagittal sinus and could be caused by reflux from the
sagittal sinus. However, these data gathered from
autopsy cases in traumatic as well as nontraumatic
cases may not reflect the pathophysiology in live
subjects with traumatic meningeal bleeding. In addition,
autopsy data give no indication on the kinetics of
meningeal bleeding. Squier and Mack concluded that
“there is a real need for detailed observational studies of
the natural history of infant SDH” [28].
In order to study the natural history of traumatic
meningeal bleeding in infants, we selected cases of infant
victims of corroborated IHI or AT who had serial CT scans.
The CT scans were studied using a semiquantitative scale
derived from the grading designed by Fisher for aneurysmal
bleeding [9].
Materials and methods
Our database was started in April 2001, following previ-
ously published methodology [35], and had since included
prospectively 469 cases of head injuries in children
under the age of 2 years. Among these, 84 were
corroborated cases of IHI or AT. We selected cases of
corroborated IHI, defined as abuse confessed by the
perpetrator, the data being provided by the judiciary to
legal physicians during expertise processes, and cases of
AT independently witnessed accidents having occurred
in public. We did not include cases of birth injury,
which represent a different group with regard to age and
pathophysiology. Corroborated cases of IHI and AT are
the subject of a companion paper. Among these, we
selected cases for which at least three CT scans
Childs Nerv Syst (2010) 26:755–762
performed during the acute phase were available for
review. The time of AT was noted from the admission
report; regarding IHI, we chose the time of clinical
onset of symptoms reported on admission as an
approximation for the time of abuse because several
reports have established that symptoms of IHI generally
occur immediately after abuse [3, 29, 30].
The first CT scan was often performed in other centers
before referral, following different protocols, including
axial views without contrast. Control CT scans were all
performed in our institution following a standard proto-
col (typically 120 kV, 215 mA, total irradiation dose
458 mGy cm). The indication for and timing of control
CT scans were dictated by the need for the medical and
surgical management of the patient.
Each CT scan was reviewed, and data were collected on
the date of the CT, volume, and distribution of bleeding and
thickness of the SDH. In order to study the extent of
bleeding on CT scan, we devised a hemorrhage score
derived from the Fisher grading widely used to evaluate
aneurysmal rupture [9]; this semiquantitative grading rates
meningeal bleeding as 0—absence of blood, 1—faint, 2—
frank, and 3—blood in thick layers or hematoma. We
defined four regions of interest (ROI) for the grading of
meningeal bleeding: the convexity of the brain (under the
calvaria, at some distance from the midline), the superior
sagittal sinus (under the calvaria close to the midline), the
falx cerebri (midline meninges at a distance from the
calvaria); and the tentorium cerebelli. The grading in these
different ROI is illustrated in Fig. 1. In addition, the
thickness of the subdural collection as noted on preopera-
tive CT scans was defined as its maximal depth, measured
perpendicular to the cortex on axial views.
These data were plotted in Figs. 2, 3, 4, 5, and 6 with the
delay from trauma to CT as the abscissa and the
Fig. 1 Grading of bleeding in the different regions of interest (ROI)
Childs Nerv Syst (2010) 26:755–762
Fig. 2 Time evolution of meningeal blood at the convexity of the
brain in AT and IHI. Each line represents an individual observation.
As a general rule, the score was maximal on the first CT scanner and
then decreased after a variable delay, depending on the initial
importance of bleeding. Note that blood at the convexity can
disappear completely within a few days after trauma
hemorrhage score or SDH thickness score as the ordinate.
In order to provide a graphic representation of the evolution
of lesions with time for each individual, each observation
was presented as a three-point plus line. All the observa-
tions in the different ROI were plotted together in the
corresponding diagrams. In order to assess the kinetics of
meningeal bleeding between the different ROI, we studied
the mean delay to the peak of hemorrhage score in the
different ROI; the delay to the peak was the time elapsed
between the trauma and the earliest CT scan showing the
highest value attained by the hemorrhage score (Fig. 7). In
order to compare the natural history of meningeal bleeding
in IHI and AT, we noted the presence of blood in the
different ROI on the first, second, and third CT scanners in
the two groups; these data, along with clinical data on
admission and at last control, are summarized in Table 1.
Means were compared using Student’s t test, using
commercially available software.
Fig. 3 Time evolution of meningeal blood at the level of the superior
sagittal sinus in AT and IHI. Each line represents an individual
observation. As a general rule, the score increased initially and then
decreased after a variable delay. The delay to peak was generally
shorter, and the delay to decrease was longer, when the initial bleeding
was more severe. Note that blood did not disappear completely along
the sagittal sinus during the period of the study
757
Fig. 4 Time evolution of meningeal blood at the level of the falx
cerebri in AT and IHI. Each line represents an individual observation.
As a general rule, the score increased initially and then decreased after
a variable delay. The delay to peak was generally shorter, and the
delay to decrease was longer, when the initial bleeding was more
severe. Note that blood can disappear completely within less than a
week after trauma
Results
We selected 20 cases of children under 24 month,
victims of confessed IHI or witnessed AT, who had at
least three CT scanners during the acute phase. The
mechanism of IHI was shaking in seven cases and
beating in three cases. AT were due to traffic accident
in eight cases and falls from windows in two cases. The
clinical and radiological features, comparing IHI and AT,
are summarized in Table 1.
The temporal variations of the hemorrhage score in
the different ROI for each individual observation are
shown in Figs. 2, 3, 4, and 5. The hemorrhage score at the
convexity of the brain was maximal on the first CT in all
cases; the score decreased at a variable speed, which
appeared faster if the initial score was low and slower if
the initial score was maximal (Fig. 2). The hemorrhage
score at the falx and sagittal sinus was generally not
Fig. 5 Time evolution of meningeal blood at the level of the
tentorium cerebri in AT and IHI. Each line represents an individual
observation. As a general rule, the score increased initially, more
slowly than for the superior sagittal sinus or the falx cerebri, and then
decreased after a variable delay. The delay to peak was also generally
shorter, and the delay to decrease was longer, when the initial bleeding
was more severe. Note that blood tends to disappear completely within
a few weeks after trauma
758
Fig. 6 Time evolution of subdural collections in AT and IHI. Each
line represents an individual observation, before drainage. The
increase in thickness was highly variable, and spontaneous regression
was witnessed in some occasions; in the majority, however, the
collections regressed only after surgical drainage
maximal from the beginning, increased with some delay, then
decreased; when the initial rate was high, the slope of the
increase was steeper, and the decrease was slower; the opposite
was found when the initial score was low (Figs. 3 and 4). The
evolution of the hemorrhage score at the tentorium was
similar to findings at the falx and sagittal sinus but with more
delay (Fig. 5). In general, the blood at the convexity
disappeared completely within a few days after trauma,
whereas the decrease occurred much later in the other ROI.
The mean delay to peak of the hemorrhage score for the
different ROI is shown in Fig. 7. The delay to the peak was
longer for the sagittal sinus and falx, compared with the
convexity, although the difference was not statistically
significant. The delay to the peak for the tentorium was,
however, significantly longer compared with the three other
sites.
The evolution of the thickness of subdural collection in
the different observations is presented in Fig. 6.
Discussion
Analysis of biases
Our study is based on a limited subset of patients because
of restrictive inclusion criteria, which introduces a selection
bias. Most patients in our prospective register were not
included because either their trauma lacked corroboration
or they underwent less than three CT scans. Inclusion in the
present series therefore implies a prolonged history with a
longer hospital stay and often an evolution toward a SDH
requiring drainage. Furthermore, both the mildest cases
(who did not require control imaging) and the most severe
(who died quickly) were not included. The vast majority of
AT occurred at home (not corroborated) or had simple
impact lesions not requiring repeated imaging. In conse-
Childs Nerv Syst (2010) 26:755–762
quence, the cases of AT in the present study group may not
be representative of the whole group; this does not diminish
the finding that the natural history of meningeal bleeding
shows little difference between IHI and AT.
AT are easily datable when these accidents occur in public
and are witnessed. Regarding IHI, precise timing is more
problematic because the perpetrator’s confession may be
absent, imprecise, or even deliberately misleading. However,
several reports have established that symptoms of IHI
generally occur immediately after abuse [3, 29, 30]; our
choice to use the time of clinical onset as an approximation
for the time of abuse was dictated by the fact that the police
investigation data at our disposal often lacked precision on
timing. In a companion paper to the present study, we found
that the delay to admission was longer for IHI than for AT,
without reaching statistical significance; however, the delay
to the first CT was significantly longer in IHI likely because
of the misleading initial presentation. In the present study,
with a more limited number of patients, the delay from
trauma to first CT was also shorter in AT; however, only the
delay to the second CT showed a statistically significant
difference. These differences may account for some of the
discrepancies between the patterns of meningeal blood in IHI
and AT as shown in Table 1.
In our study, the initial imaging was often made in different
hospitals, following similar protocols of standard use for head
injuries in infants (plain CT, axial view, brain window). We do
not think the technique of the CT scan impacted much on our
analysis of the extent of meningeal bleeding. Further CT scans
were made on request, according to the need for medical
management of the child; these variations in timing imposed
on us the diagram presentation in Figs. 2, 3, 4, 5, and 6.
Fig. 7 Mean delay between the trauma and the peak of meningeal
blood in the different ROI. For the convexity of the brain, this delay to
the peak corresponded to the delay to the first CT scanner. The
difference between delay to the peak at the convexity and along the
sagittal sinus and the falx did not reach statistical significance;
however, the delay to peak at the tentorium was statistically longer.
Our interpretation is that the meninges at the convexity of the brain are
the site of original bleeding, after which part of the blood migrates to
the midline following the circulation and absorption of CSF (falx and
sagittal sinus), while another part sediments following gravity to the
recumbent regions of the skull (tentorium)
Childs Nerv Syst (2010) 26:755–762 759

Role of CT Table 1 (continued)

Abuse Accident p value

Since its introduction, CT scanning has been little exploited
as a source of information on head injuries in infants Mean duration of stay
probably because MRI soon emerged as a superior tool for Under sedation (days) 4.9 1.3 0.02
the study of brain lesions, dating of trauma, and diagnosis In ICU (days) 7.2 2.6 0.02
of repeated assault [7, 26]. However, CT scanning has In hospital (days) 21.0 12.4 0.004
Outcome at last control
Table 1 Comparison of clinical and radiological data in AT and IHI Age (months) 26.1 25.7 NS
Sequels 4 1
Abuse Accident p value
Mortality 1 0
Patients
Only means were compared, using Student’s t test; no chi-square test
N 10 10
could be performed because the theoretical values were all under 5. In
M/F 6/4 6/4 this table, only the presence or absence of blood in the different ROI is
Mean age (months) 4.1 7.4 NS reported, with no attempt at quantifying it. This table shows that AT
Signs and symptoms at presentation tended to have more often signs of impact and/or skull fracture than
IHI and that meningeal blood tended to be more widespread in the IHI
Signs of impact 3 10 group, correlated with a more severe clinical presentation and
Coma 3 3 outcome. However, the time-related pattern of blood distribution
Epilepsy 9 3 between the different ROI appeared to depend more on the delay after
trauma than on its cause. For example, the absence of blood along the
Status epilepticus 3 0
tentorium on the second CT scanner in several cases of AT may be
Raised ICP 10 4 related to the significantly shorter delay to second CT scanner in this
First CT group. In summary, the patterns of AT and IHI differed by the
Mean delay (hours) 17.7 8.7 NS presence or absence of signs of impact and the extent of meningeal
bleeding; however, the natural history of meningeal hemorrhage in the
Blood two groups did not appear different
Convexity 10 7
Sagittal sinus 9 9
Falx 10 8
remained of value because of its availability, repeatability,
Tentorium 5 4
and accuracy to diagnose small amounts of fresh blood that
Subdural
would be undetected by MRI [26, 31]. The ease with which
Thickness (mm) 4.2 2.7 NS
CT scan can be repeated allows observations on the kinetics
Second CT
of the lesions [7]. However, no systematic study on the
Mean delay (days) 2.7 1.5 0.04
natural history of meningeal bleeding with imaging has
Blood
been performed probably because of the scarcity of accurate
Convexity 8 6 data, of the difficulty to establish clinicoradiological
Sagittal sinus 10 10 correlates, and of the lack of grading system. In an earlier
Falx 9 9 study, we noted that meningeal blood clots appeared to
Tentorium 10 4 change location between one study and the next [34]. We
Subdural decided to focus on this unreported phenomenon because it
Thickness (mm) 4.9 2.9 NS appeared to illustrate and complement pathological find-
Third CT ings, such as thickening of the falx [16, 36], infiltration of
Mean delay (days) 7.1 12.5 NS the dural channels by red blood cells, and absence of
Blood tearing of the corticodural veins [12, 28]. For our present
Convexity 7 5 study, we selected precisely dated trauma, in order to have a
Sagittal sinus 8 9 reliable starting point, and repeated observations with
Falx 9 7 imaging. For these reasons, we restricted our dataset to
Tentorium 9 9 corroborated IHI or AT, with at least three CT scans.
Subdural
Thickness (mm) 6.1 2.2 0.008 CT evaluation
Raised ICP
Mean delay (days) 1.9 6.1 0.016 We did not find in the literature a scale that quantita-
Puncture 8 6 tively evaluated meningeal bleeding in infantile traumatic
Drainage 6 3 injuries with the aim of studying its natural history. We
760
chose to adapt Fisher’s grading [9] because it has been
validated and accepted for a long time, its four-tier scale
has proved both informative and reliable, and it has
spawned a number of variant evaluation scales. We
defined the different ROI based on our previous experi-
ence and on the focus of the literature on the sagittal sinus
[10, 19, 27] and falx cerebri [7, 16, 36]. Considering the
results in our study, we propose that this scale system
could be used for the evaluation of the natural history of
traumatic head injuries in infants.
Pathophysiology
The origin of meningeal bleeding in IHI has long been
thought to be the tearing of corticodural bridging veins
along the sagittal sinus [16, 21, 23, 36]. However,
autopsy demonstrated an absence of tearing of these veins
in many cases [12, 19, 28]. Tearing of the bridging veins
can sometimes be found during craniotomy for massive,
clotted SDH; however, it appears to be a marginal cause of
meningeal bleeding in infants. In addition, the energy
required to tear the veins is considered higher than that
resulting from abusive shaking [23, 25]. The origin of
meningeal bleeding is now seen by several authors in the
dura mater itself [5, 19, 28]. In sophisticated micro-
anatomical studies, several authors have described venous
plexuses, as well as dural “holes” in the vicinity of the sagittal
sinus, which are likely involved in the absorption of
cerebrospinal fluid (CSF) in infants whose arachnoid villi
are immature [10, 19, 27]. Pointing out the characteristic
thickening of the falx and paramedian location of blood in
infantile SDH, several authors postulated that the meningeal
bleeding could be the result of oozing from these plexuses
[5, 19, 28]. Our data appear in accordance with the hypothesis
of bleeding from the dura mater but suggest instead that the
initial bleeding occurs at some distance from the midline and
that the thickening of the falx is a secondary feature.
The traumatic nature of meningeal bleeding in
infantile SDH has been challenged in recent years. The
emphasis was put on “alternative source of bleeding in
nontraumatic condition” [19] and on hypoxic mechanisms
causing brain damage as well as meningeal bleeding [13],
spurring heated controversy considering the legal impli-
cations [4]. Squier and Mack formulated the hypothesis of
a reflux from the venous sinus to the dural venous
plexuses close to the superior sagittal sinus [19, 28]. This
hypothesis was based on the autopsy finding of dural
bleeding in the paramedian regions, in patients without
evidence of impact, and even in patients who died of
nontraumatic causes. Whatever the origin of bleeding and
brain damage, we think that the traumatic nature of the
lesions in our series is beyond doubt. In cases of IHI, we
believe that the primum mobile of lesions remains
Childs Nerv Syst (2010) 26:755–762
forceful, abusive shaking, as attested by the delay to
referral, the absence of spontaneous confession, the
context of neglect and familial dysfunction, and confirmed
by confession from the perpetrators [3, 18, 29, 30]. We
also found that the pattern of lesions showed little
dissimilarity between IHI and AT but that the pathophys-
iology appeared quite similar in both groups. This
suggests that the natural history of meningeal bleeding is
related to age rather than to the cause of trauma. One may
wonder whether the autopsy findings described in case of
traumatic as well as nontraumatic death [28] are identical
to our findings in corroborated trauma cases or are the
result of other, nonspecific, per-mortem processes. Post-
mortem imaging before autopsy may become necessary to
answer these questions [15].
The evolution of traumatic lesions with time gives
indications for the dating of trauma in legal cases. Most
publications have focused on the modifications of blood
density or MRI signal (which can be misleading);
however, several authors have also reported redistribu-
tion of blood from the convexity to the falx [7, 17, 34] or
sedimentation to the tentorium and the spinal axis [5, 8,
28]. Unlike autopsy, imaging, especially CT scan, can be
repeated at several stages of the disease, allowing the
follow-up of time-related alterations [7]. In the absence of
drainage, the spontaneous evolution toward chronic SDH,
with thickening of fibrous membranes and rebleeding, is a
classical outcome in adults and also occurs in some infants
with brain atrophy [11]. In most infants, however, the
SDH evolves quite differently, the accumulation of CSF
playing a major role in its pathophysiology [32]. Our
hypothesis is that after the initial bleeding, blood cells
migrate, following the gradient of circulation of CSF
toward its absorption sites, likely the dural “holes”
described by Squier et al. [27]. Histologically, red blood
cells have been assimilated to “natural tracers of CSF
absorption” [38]. We think this is no coincidence that
blood accumulates in the meningeal sites that are
responsible for the absorption of CSF, at the same time
CSF accumulates in the subdural space. More research is
needed to confirm this hypothesis.
Spontaneous disappearance of acute SDH has also been
reported [6, 17, 20]. The delay to disappearance of subdural
blood in neonates has been established at less than 4 weeks
[24, 37]; these findings have important medicolegal
implications since birth injury is sometimes presented as a
cause of unexplained SDH diagnosed later in life. In our
study, the time to disappearance of meningeal blood varied
following the site and the initial importance of bleeding but
could be as quick as a few days at the convexity. This
emphasizes the importance of careful evaluation of CT
scanners for the presence of residual bleeding in sites like
the midline and posterior fossa [7].
Childs Nerv Syst (2010) 26:755–762
Conclusion
In the present study, we studied CT scanner following a
semiquantitative method for the study of meningeal
bleeding in different ROI. Using this method on a selected
group of patients with known time of trauma, this method
allowed us to approach the natural history of meningeal
bleeding. We found that blood was maximal from the
beginning at the convexity and increased with delay on the
midline sites, then along the tentorium cerebelli. These
findings lead us to formulate hypotheses on the site of
initial bleeding and the relation between meningeal bleed-
ing and clogging of the CSF absorption pathways leading
to CSF accumulation in the subdural space. Knowledge of
these time-related modifications may be of interest for the
timing of traumatic lesions and the diagnosis of repeated
trauma in legal cases. These findings need to be confirmed
and validated with more cases.
References
1. Barkovich AJ (2000) Brain and spine injuries in infancy and
childhood. In: Barkovich AJ (ed) Pediatric neuroradiology, 3rd
edn. Lippincott Williams and Wilkins, Philadelphia, pp 157–249
2. Barlow KM, Gibson RJ, McPhillips M, Minns RA (1999)
Magnetic resonance imaging in acute non-accidental head injury.
Acta Paediatr 88:734–740
3. Biron D, Shelton D (2005) Perpetrator accounts in infant abusive
head trauma brought about by a shaking event. Child Abuse
Neglect 29:1347–1358
4. Byard RW, Blumbergs P, Rutty G, Spehake J, Banner J, Krous HF
(2007) Lack of evidence for a causal relationship between
hypoxic–ischemic encephalopathy and subdural hemorrhage in
fetal life, infancy, and early childhood. Ped Developmental
Neuropathol 10:348–350
5. Case ME (2008) Inflicted traumatic injury in infants and young
children. Brain Pathol 18:571–582
6. Duhaime AC, Christian C, Armonda R, Hunter J, Hertle R (1996)
Disappearing subdural hematomas in children. Pediatr Neurosurg
25:116–122
7. Feldman KW, Brewer DK, Shaw DW (1995) Evolution of the
cranial computed tomography scan in child abuse. Child Abuse &
Neglect 19:307–314
8. Feldman KW, Weinberger E, Milstein JM, Fligner CL (1997)
Cervical spine MRI in abused infants. Child abuse Neglect
21:199–205
9. Fisher CM, Kistler JP, Davis JM (1980) Relation of cerebral
vasospasm to subarachnoid hemorrhage visualized by computer-
ized tomographic scanning. Neurosurgery 6:1–9
10. Fox RJ, Walji AH, Mielke B, Petruk K, Aronyk KE (1996)
Anatomic details of intradural channels in the parasagittal dura: a
possible pathway for flow of cerebrospinal fluid. Neurosurgery
39:84–91
11. Friede RL (1989) Subdural hematomas, hygromas, and effusions.
In: Friede RL (ed) Developmental neuropathology. Springer,
Berlin, pp 198–208
12. Geddes JF, Hackshaw AK, Vowles GH, Whitwell HL (2001)
Neuropathology of inflicted head injury in children. I. Patterns of
brain damage. Brain 124:1290–1298
761
13. Geddes JF, Tasker RC, Hackshaw AK, Nickols CD, Adams GGW,
Whitwell HL, Scheimberg I (2003) Dural hemorrhage in non-
traumatic infant death: does it explain the bleeding in ‘shaken
baby syndrome’? Neuropathol Appl Neurobiol 29:14–22
14. Guthkelch AN (1971) Infantile subdural haematoma and its
relationship to whiplash injuries. BMJ 2:430–431
15. Hart BL, Dudley MH, Zumvalt RE (1996) Postmortem cranial
MRI and autopsy correlation in suspected child abuse. Am J
Forensic Med Pathol 17:217–224
16. Harwood-Nash DC (1992) Abuse to the pediatric central nervous
system. Am J Neuroradiol 13:569–575
17. Kuroiwa T, Tanabe H, Takatsuka H, Arai M, Sakai N, Nagasawa
S, Ohta T (1993) Rapid spontaneous resolution of acute extradural
and subdural hematomas. Case report. J Neurosurg 78:126–128
18. Leestma JE (2006) “Shaken baby syndrome”: do confessions by
alleged perpetrators validate the concept? J Am Phys Surg 11:14–16
19. Mack J, Squier W, Eastman JT (2009) Anatomy and development
of the meninges: implications for subdural collections and CSF
circulation. Pediatr Radiol 39:200–210
20. Modic MT, Kaufman B, Bonstelie CT, Tomsick TA, Weinstein
MA (1981) Megalocephaly and hypodense extracerebral fluid
collections. Radiology 141:93–100
21. Papasian NC, Frim DM (2000) A theoretical model of benign
external hydrocephalus that predicts a predisposition toward extra-
axial hemorrhage after minor head trauma. Pediatr Neurosurg
33:188–193
22. Prange MT, Coats B, Duhaime AC, Margulies SS (2003)
Anthropomorphic simulations of falls, shakes, and inflicted
impacts in infants. J Neurosurg 99:143–150
23. Raul JS, Roth S, Ludes B, Willinger R (2008) Influence of the
benign enlargement of the subarachnoid space on the bridging
veins strain during a shaking event: a finite element study. Int J
Legal Med 122:327–340
24. Rooks VJ, Eaton JP, Ruess L, Petermann GW, Keck-Wherley J,
Pedersen RC (2008) Prevalence and evolution of intracranial
hemorrhage in asymptomatic term infants. Am J Neuroradiol
29:1082–1089
25. Roth S, Raul JS, Ludes B, Willinger R (2007) Finite element
analysis of impact and shaking inflicted to a child. Int J Legal
Med 121:223–228
26. Sato Y, Yuh WTC, Smith WL, Alexander RC, Kao SCS,
Ellerbroek CJ (1989) Head injury in child abuse: evaluation with
MR imaging. Radiology 173:653–657
27. Squier W, Lindberg E, Mack J, Darby S (2009) Demonstration of
fluid channels in human dura and their relationship to age and
intradural bleeding. Child’s Nerv Syst 25:925–931
28. Squier W, Mack J (2009) The neuropathology of infant subdural
haemorrhage. Forensic Sci Int 187:6–13
29. Starling SP, Holden JR, Jenny C (1995) Abusive head trauma:
the relationship of perpetrators to their victims. Pediatrics
95:259–262
30. Starling SP, Patel S, Burke BL, Sirotnak AP, Stronks S, Rosquist
P (2004) Analysis of perpetrator admissions to inflicted traumatic
brain injury in children. Arch Pediatr Adolesc Med 158:454–458
31. Stoodley N (2002) Non-accidental head injury in children:
gathering the evidence. Lancet 360:271–272
32. Vinchon M, Noulé N, Soto-Ares G, Dhellemmes P (2001)
Subduroperitoneal drainage for subdural hematomas in infants:
results in 244 cases. J Neurosurg 95:249–255
33. Vinchon M, Noizet O, Defoort-Dhellemmes S, Soto-Ares G,
Dhellemmes P (2002) Infantile subdural hematomas due to traffic
accidents. Pediatr Neurosurg 37:245–253
34. Vinchon M, Noulé N, Tchofo PJ, Soto-Ares G, Fourier C,
Dhellemmes P (2004) Imaging of head injuries in infants:
temporal correlates and implications for the diagnosis of child
abuse. J Neurosurg 101(1 Suppl):44–52
762
35. Vinchon M, Defoort-Dhellemmes S, Desurmont M, Dhellemmes
P (2005) Accidental and nonaccidental head injuries in infants: a
prospective study. J Neurosurg (Pediatrics 4) 102:380–384
36. Wells RG, Vetter C, Laud P (2002) Intracranial hemorrhage in
children younger than 3 years: prediction of intent. Arch Pediatr
Adolesc Med 156:252–257
Childs Nerv Syst (2010) 26:755–762
37. Whitby EH, Griffiths PD, Rutter S, Smith MF, Sprigg A, Ohadike
P, Davies NP, Rigby S, Paley MN (2004) Frequency and natural
history of subdural haemorrhages in babies and relation to
obstetric factors. Lancet 363:846–851
38. Yamashima T (1986) Ultrastructural study of the final cerebrospinal
fluid pathway in human arachnoid villi. Brain Res 384:68–76