1. Sonographic Visualization of
Neonatal Posterior Fossa
Abnormalities Through the
Posterolateral Fontanelle
Joseph A. Luna 1 OBJECTIVE. This study was performed to determine whether imaging through the poste-
Ruth B. Goldstein 2 rolateral fontanelle in addition to the anterior fontanelle during neonatal cranial sonography
improves diagnostic accuracy or examiner confidence in the diagnosis of neonatal posterior
fossa abnormalities.
MATERIALS AND METHODS. In 1995 we changed our protocol of neonatal cranial
sonography to include imaging through the posterolateral fontanelle in all patients. The re-
ports of all sonography performed in the first 15 months of this protocol were reviewed, and
two radiologists reviewed the images of all patients in whom a posterior fossa abnormality
was diagnosed with posterolateral fontanelle images masked and then with posterolateral fon-
tanelle images available.
RESULTS. In total, 1292 sonograms were obtained in 462 patients. In 200 patients, the
sonographic findings were abnormal; of these 200 patients, 24 (12%) had posterior fossa ab-
normalities (nine posterior fossa hemorrhages, four Arnold-Chiari malformations (type II),
two posterior fossa arteriovenous malformations, and nine partial vermian defects). The pos-
terolateral fontanelle images showed the posterior fossa abnormality better than the anterior
fontanelle images did in 23 (96%) of the 24 patients, increased confidence in the diagnosis of
18 (75%) of the 24 patients, and was the only technique to reveal the posterior fossa abnor-
mality in 11 (46%) of the 24 patients. Nearly all pathologic correlations with imaging con-
firmed the posterolateral fontanelle findings except for the diagnosis of inferior vermian
agenesis, which was presumed to be a false-positive diagnosis in four patients in whom MR
imaging showed no abnormalities.
CONCLUSION. Additional imaging through the posterolateral fontanelle during routine
neonatal cranial sonography added considerable benefit. False-positive diagnosis of vermian
defects is a troubling problem but may be avoided with careful attention to the midline sagittal
sonographic images of the vermis and fourth ventricle.
S
onography of the neonatal brain is an conventional anterior fontanelle images obtained
important tool in the assessment of in all cranial sonography performed at our insti-
neonates, particularly premature ne- tution. The purpose of this review of our first 15
onates at significant risk for intracranial hemor- months’ experience is to determine the benefit of
rhage. The procedure has shown good sensitivity adding posterolateral fontanelle imaging to the
and specificity in the detection of neonatal intra- conventional anterior fontanelle examination.
cranial abnormalities, particularly in the suprat-
entorial region [1–3]. When scanning through the Materials and Methods
anterior fontanelle, the most poorly evaluated re- The studies were performed in an academic refer-
Received May 14, 1999; accepted after revision gion is the posterior fossa. This is largely because ral center with a 42-bed regional intensive care unit
July 22, 1999. approved for extracorporeal membrane oxygenation
the posterior fossa is farthest from the transducer
1
Kaiser Permanente, 4647 Zion Ave., San Diego, CA 92120. (ECMO) and pediatric cardiothoracic surgery.
and because many of its structures are parallel to
2 We reviewed the reports of all of the cranial sono-
Department of Radiology, University of California, the insonating beam [4–7]. Several articles have
505 Parnassus Ave., M-396, San Francisco, CA 94143-0628. grams obtained during the first 15 months of using
recommended adjunctive imaging through poste- routine posterolateral fontanelle sonography in addi-
Address correspondence to R. B. Goldstein.
rior and posterolateral fontanelles to improve vi- tion to our standard cranial sonography (February
AJR 2000;174:561–567
sualization of the posterior fossa [4–8]. 1995 through April 1996). Each patient whose sono-
0361–803X/00/1742–561 In 1995 we added posterolateral fontanelle graphic report indicated a posterior fossa abnormal-
© American Roentgen Ray Society angled axial images of the cerebellum to the ity was identified.
AJR:174, February 2000 561

2. Luna and Goldstein
A B C
Fig. 1.—5-day-old premature male neonate born at 25 weeks’ gestation.
A, Sonogram obtained through posterolateral fontanelle shows normal cerebellar hemispheres (c), echogenic cerebellar vermis (curved arrow), and fourth ventricle (straight
arrow). L = left.
B, Sonogram obtained 3 days later through anterior fontanelle approach shows possible left hemorrhage (arrow).
C, Posterolateral fontanelle image obtained on same day as B shows unequivocal left cerebellar hemorrhage (arrows). lt = left, rt = right.
Scanning was performed portably in the neonatal ence or absence of a posterior fossa abnormality; and than 32 weeks’ gestational age at birth. Birth
intensive care unit by experienced sonographers using when a posterior fossa abnormality was present, the weights in the neonates with posterior fossa ab-
phased array multiple foci 5.0- or 7.0-MHz sector examiner was asked to make a specific diagnosis if normalities ranged from 915 to 4000 g.
transducers (XP; Acuson, Mountain View, CA). Dur- possible (e.g., cerebellar hemorrhage, intraventricular
ing each cranial sonogram six to eight images were ob- or extraaxial hemorrhage, ventricular dilatation, ver- Posterior Fossa Abnormalities
tained through the anterior fontanelle in both the mian agenesis, or other abnormality). Next, the poste- Of the 24 patients with posterior fossa abnor-
coronal and parasagittal planes for a total of 14–16 im- rolateral fontanelle images were unmasked and the malities, nine had posterior fossa hemorrhage:
ages, including images of the posterior fossa. After this entire sonogram was rereviewed independently by six had isolated posterior fossa hemorrhage, one
imaging, additional focused images of the posterior two observers. Again, the presence or absence of pos-
had hemorrhage as a result of posterior fossa du-
fossa were obtained through the posterolateral fonta- terior fossa abnormality was noted and a specific
nelle. The transducer was placed over the postero- diagnosis was sought. Finally, with all the images
ral arteriovenous malformation, and two had
lateral fontanelle (using the most accessible upside available, each examiner was asked to make an inde- fourth ventricular and extraaxial blood associated
fontanelle) by gently bending the superior portion of pendent subjective assessment of which images better with marked supratentorial subependymal hem-
the pinna down to find the acoustic window of the pos- displayed the posterior fossa abnormality and whether orrhage (SEH) and intraventricular hemorrhage
terolateral fontanelle. The transducer was then rotated adding the posterolateral fontanelle images enhanced (grade III SEH). Of the remaining patients, four
slightly using real-time imaging to obtain the appropri- the confidence with which he or she made the diagno- had findings related to the Arnold-Chiari malfor-
ate angled axial image depicting the cisterna magna, sis of posterior fossa abnormality. mation II (small posterior fossa, effaced cisterna
cerebellar hemispheres, vermis cerebelli, fourth ventri- Clinical follow-up was obtained through discus- magna), two had posterior fossa arteriovenous
cle, tectal lamina, and pons. Four to six images were ob- sion with the neonatologists, review of clinical charts, malformations without hemorrhage, and nine
tained in this plane from superior to inferior, generally study of any additional imaging, and review of avail-
neonates were thought to have partial (inferior)
imaging from the incisura tentorii cerebelli through the able autopsy results to determine if findings on the
inferior vermis. A normal cerebellum imaged through posterolateral fontanelle images significantly altered
vermian defects, or inferior vermian agenesis.
the posterolateral fontanelle shows the biconvex dorsal the clinical diagnosis or prompted additional imaging Images of the Lesion
margin of the cerebellar hemispheres, normal cerebellar studies on the neonate (including sonography, MR
The examiners thought the posterolateral fon-
vermis, and the fourth ventricle (Fig. 1A). imaging, or CT). Confirmatory studies or follow-up
tanelle images displayed the posterior fossa ab-
Almost all neonates were imaged through the for each patient were noted when available.
posterolateral fontanelle during this study period. normality better than the anterior fontanelle
The only exceptions were some neonates on ECMO images in 23 of the 24 patients. The single abnor-
(vascular access catheter placement would not allow
Results mality better seen with anterior fontanelle images
slight turning of the neck). The precise number of During the 15-month review period, 1292 cra- was an extremely small and distorted posterior
ECMO neonates not examined through the postero- nial sonograms were obtained in 462 patients. fossa related to an Arnold-Chiari malformation
lateral fontanelle could not be determined from this Gestational ages at birth in this patient population II. The severely distorted posterior fossa ob-
retrospective review. No patient in our study with a ranged from 24 to 42 weeks. In 262 patients, the scured the usual posterior fossa landmarks, and
recognized posterior fossa abnormality was exam- cranial sonographic findings were reported to be the conventional anterior fontanelle images were
ined only through the anterior fontanelle. normal and no further follow-up was obtained. In more easily interpreted in this patient.
The cranial sonograms of the patients with poste- 200 patients, the sonographic findings were inter- The posterolateral fontanelle images in-
rior fossa abnormalities were independently reviewed
preted to be abnormal, and 24 patients (12%) had creased confidence in the diagnosis of abnor-
by two radiologists, unaware of clinical diagnosis, in
the following fashion: first, with the posterolateral
sonographic abnormalities of the posterior fossa. malities in 18 (75%) of the 24 unhealthy
fontanelle images masked, the conventional coronal Of these 24 patients, 11 were term neonates and patients (Figs. 1–3). These 18 neonates in-
and parasagittal images obtained through the anterior 13 were preterm (<36 weeks’ gestational age at cluded eight of the nine patients with posterior
fontanelle were reviewed. Note was made of the pres- birth). Eight (33%) of the 24 neonates were less fossa hemorrhage (Figs. 1 and 2), including
562 AJR:174, February 2000

3. Sonography of Neonatal Posterior Fossa Abnormalities
Fig. 2.—3-day-old premature female
neonate born at 27 weeks’ gestation.
Posterolateral fontanelle images
more sensitively detect cerebellar
hemorrhage.
A, Anterior fontanelle images do not
show posterior fossa hemorrhage.
R = right.
B, Posterolateral fontanelle images
in same patient, same day, show
unequivocal echogenic unilateral
hemorrhage (arrow) in cerebellum.
L = left, R = right.
A B
A B
Fig. 3.—8-day-old male neonate in whom cerebellar hemorrhage is better seen on posterolateral fontanelle images.
A, Coronal image obtained through anterior fontanelle fails to show cerebellar abnormality (arrow).
B, Midline sagittal image obtained through anterior fontanelle shows questionable cerebellar abnormality (arrow).
C, Posterolateral fontanelle image clearly shows bilateral cerebellar hemorrhages (arrows). Neonate died of
complications of necrotizing enterocolitis and did not undergo either autopsy or MR imaging.
C
AJR:174, February 2000 563

4. Luna and Goldstein
both patients with fourth ventricle hemor- six of the 11 cases: two of the three posterior topsy (Fig. 3). The two patients with posterior
rhage associated with Grade III SEH (Fig. fossa hemorrhages were confirmed with au- fossa dural arteriovenous malformations had
4), one patient with Arnold-Chiari malforma- topsy or MR imaging; the other was monitored confirmed diagnoses with MR imaging. One
tion II, and all nine neonates with suspected with serial sonograms showing evolution and patient died at 13 days of life from complica-
inferior vermian agenesis (Figs. 5 and 6). gradual resolution of the lesion. In two cases tions related to a large myelomeningocele and
In five (21%) of the 24 patients, although (posterior fossa extraaxial and intraventricular did not undergo MR imaging. Nine neonates
the abnormality was better seen on the poster- hemorrhage) the abnormalities were not con- were thought to have inferior vermian agene-
olateral fontanelle images, the improved vi- firmed with additional imaging but were associ- sis as a result of an unusually large communi-
sualization was not considered to increase ated with large supraventricular subependymal cation between the fourth ventricle and the
examiner confidence in diagnosis because and intraventricular hemorrhages (grade III cisterna magna. MR imaging was performed
the lesion was already adequately displayed SEH) (Fig. 4). Of the six cases of possible or in six of these neonates and showed inferior
by the anterior fontanelle images. These five probable inferior vermian agenesis detected vermian agenesis in two neonates and normal
abnormalities included both posterior fossa only on the posterolateral fontanelle images cerebellum in the other four. Neither MR im-
arteriovenous malformations, two small pos- (Figs. 5 and 6), four were evaluated with MR age nor autopsy was performed in the other
terior fossae related to the Arnold-Chiari imaging. Three were shown to have a normal three neonates thought to have inferior ver-
malformation II, and one 5-cm posterior vermis cerebelli (false-positive sonographic di- mian agenesis.
fossa hemorrhage. In one of the 24 patients, agnoses) on MR imaging and one showed infe-
the abnormality of the posterior fossa was rior vermian agenesis. Altered Diagnosis
better visualized using conventional anterior In total, follow-up MR images, autopsy, or Adding the posterolateral fontanelle im-
fontanelle imaging. follow-up sonography was available for 16 of ages to our standard cranial sonograms al-
In 11 (46%) of the 24 patients, one or both the 24 unhealthy neonates. Six neonates with tered diagnosis or prompted additional
examiners saw significant abnormalities exclu- posterior fossa hemorrhage underwent confir- evaluation (including imaging studies) in 12
sively on the posterolateral fontanelle images. matory studies (MR imaging, three; autopsy, of the 24 patients identified with posterior
These abnormalities included three cases of cer- one; and serial sonograms, two). One neonate fossa abnormalities. Three posterior fossa
ebellar hemorrhage, both cases of fourth ven- with posterior fossa hemorrhage died of hemorrhages and five suspected inferior ver-
tricular and extraaxial hemorrhage, and six of complications of necrotizing enterocolitis and mian ageneses were visible only on the pos-
the nine cases of suspected inferior vermian disseminated intravascular coagulation and terolateral fontanelle images. In three other
agenesis. Confirmatory studies were obtained in did not undergo either MR imaging or au- posterior fossa hemorrhages and one sus-
A B
Fig. 4.—Large intraventricular and extraaxial hemorrhage in 7-day-old premature male neonate born at 32 weeks’ gestation.
A, Posterolateral fontanelle image shows extraaxial hemorrhage (thick arrow and cursors) and clot in fourth ventricle (thin arrow).
B, Anterior fontanelle image shows neither of the findings in posterior fossa (arrow ) seen in A.
564 AJR:174, February 2000

5. Sonography of Neonatal Posterior Fossa Abnormalities
A B
Fig. 5.—False-positive sonographic diagnosis of inferior vermian agenesis in 4-day-old male neonate.
A, False-positive inferior vermian agenesis (arrow) shown on posterolateral fontanelle image.
B, Midline sagittal image obtained through anterior fontanelle shows intact vermis (curved arrows), retrospectively appreciated, covering fourth ventricle (straight arrow).
MR imaging showed normal cerebellum.
A B
Fig. 6.—1-day-old male neonate with inferior vermian agenesis.
A, Posterolateral fontanelle image shows inferior vermian defect (arrow) that was associated with agenesis of corpus callosum (not shown).
B, Midline sagittal anterior fontanelle image reveals possible inferior vermian agenesis. Note poor covering of fourth ventricle (long arrow) by vermis (curved arrows). Also note
that partial voluming of small part of contralateral cerebellar hemisphere (small arrow) can be mistaken for intact vermis. Inferior vermian defect was confirmed on MR imaging.
AJR:174, February 2000 565

6. Luna and Goldstein
pected inferior vermian agenesis the conven- abnormalities. In 23 (96%) of the 24 sono- a small apparent communication between the
tional images showed an equivocal area that graphically diagnosed posterior fossa ab- fourth ventricle and the cisterna magna that
was interpreted as definitely abnormal on the normalities, two examiners independently was mistakenly thought to represent an infe-
posterolateral fontanelle images. concluded that the abnormality was better seen rior vermian defect. The erroneous diagnoses
using the posterolateral fontanelle images. were all made early in our study. This imag-
Interobserver Agreement Even if the posterior fossa abnormality could ing pitfall has been described in the fetal
Good interobserver agreement was seen. The be identified on the anterior fontanelle images, sonography literature (angled axial and coro-
examiners agreed with the diagnosis and “im- the examiners in this study concluded that add- nal images of the posterior fossa similar to
provement” or “no improvement” of the postero- ing the posterolateral fontanelle images re- our posterolateral fontanelle images are eas-
lateral fontanelle image for the diagnosis in 23 of sulted in increased confidence in diagnosis in ily and commonly obtained) [15]. Angled ax-
the 24 neonates. The examiners differed in the 18 (75%) of the 24 cases. ial images of the cerebellum can make a
confidence of the findings on the anterior fon- In 11 (46%) of the 24 patients, the postero- prominent vallecula appear similar to a par-
tanelle images in only one patient. One examiner lateral fontanelle approach allowed detection tial (inferior) vermian defect. Barkovich et
interpreted the images as having normal findings; of significant findings that were not clearly al. [16], who used MR imaging to investigate
the other interpreted the image as showing a seen on anterior fontanelle imaging, especially Dandy-Walker syndrome, found that the nor-
“possible abnormal area.” Both examiners inter- in the seven neonates with posterior fossa mally formed vermis may tilt forward from
preted this patient’s posterolateral fontanelle im- hemorrhage. One hemorrhage was well seen its usual position and create the impression
ages as showing posterior fossa hemorrhage. with anterior fontanelle imaging. The other six of an inferior vermian defect on angled axial
were poorly seen or not detected at all using and coronal images. We believe that scrutiny
anterior fontanelle imaging, yet confidently of midline sagittal images of the vermis cere-
Discussion identified using posterolateral fontanelle imag- belli obtained through the anterior fontanelle
Cranial sonography of the neonate is a ing. Six of these hemorrhages were confirmed may serve to arbitrate in equivocal cases
widely accepted technique for evaluating the by autopsy (n = 1), MR imaging (n = 3), or fol- (Figs. 5B and 6B). A midline sagittal image
neonatal brain. Initial reports describing the low-up sonography ( n = 2). There were no allows display of the precise cerebellar anat-
technique stressed imaging through the anterior false-positive diagnoses of posterior fossa hem- omy because the nodulus of the vermis cov-
fontanelle [9, 10]. Imaging through the anterior orrhage. Only three of the nine neonates with ering the inferior roof of the fourth ventricle
fontanelle allows excellent evaluation of the posterior fossa hemorrhage had died at the time can be better seen, indicating whether the
common sites of germinal matrix hemorrhage of this writing. Whereas large and catastrophic inferior vermis is intact or deficient. This ex-
and the cerebral ventricles, but a weakness of hemorrhages have generally been described in perience emphasizes the caution with
anterior fontanelle imaging is its evaluation of the literature, posterolateral fontanelle imaging which the diagnosis of inferior vermian
the posterior fossa [4, 5, 7]. Taylor et al. [2] and allowed the detection of some relatively small agenesis should be made.
Babcock et al. [3] described neonates in whom cerebellar hemorrhages in this study. Two large In conclusion, ours is a retrospective study,
significant posterior fossa hemorrhage was autopsy studies of low-birth-weight premature and the overall sensitivity of the posterolateral
missed on sonography performed exclusively neonates have reported an incidence of posterior fontanelle images for posterior fossa abnor-
through the anterior fontanelle. fossa hemorrhage between 16% and 21% even malities is not evaluated. However, in our re-
It has recently been hypothesized that bring- though this hemorrhage is less common than view of 15 months of experience with these
ing the high-frequency transducer closer and supraventricular and intraventricular hemor- images, the posterolateral fontanelle images
more perpendicular to many of the posterior rhage [12, 13]. In one of these studies [13] the revealed three posterior fossa hemorrhages not
fossa structures should improve image clarity. findings of cerebellar hemorrhage fell into two detected with standard anterior fontanelle im-
Several recent studies have confirmed im- groups: large hemorrhages destroying one third aging and confirmed three more that were only
proved visualization of the normal posterior or more of the cerebellar parenchyma and small suspected on the standard images. Nearly ev-
fossa using imaging through a variety of poste- hemorrhages no larger than 5 mm. ery abnormality observed was considered bet-
rior acoustic windows, including the foramen The observation of small cerebellar hem- ter displayed on the posterolateral fontanelle
magnum [4], the posterior fontanelle [5, 7], orrhages using the posterolateral fontanelle images, and adding the posterolateral fonta-
and the posterolateral fontanelle [8, 11]. approach in premature neonates has recently nelle images allowed increased confidence in
The technique for obtaining images been described [14]. The clinical signifi- diagnosis in 75% of the posterior fossa
through the posterolateral fontanelle is cance of these otherwise undiagnosed hem- abnormalities. Only 1 or 2 min of additional
learned easily and quickly. After only a few orrhages is currently under investigation. scan time is needed to obtain these images.
“learning cases” the images may be obtained Unfortunately, better visualization of The potential pitfall of overdiagnosing inferior
in less than 5 min, and our sonographers con- structures not well visualized in the past can vermian agenesis should be anticipated and
sistently obtain the images in less than 2 min, lead to false-positive diagnoses when one easily avoided.
usually in less than 1 min. embarks on using this new technique. In this
On the basis of these early reports and our study at least four false-positive diagnoses
own experience, we added routine posterolat- occurred as a result of adding the posterolat- References
eral fontanelle imaging to our standard cranial eral fontanelle images; all were misdiag-
1. Taylor GA, Fitz CR, Kapur S, Short BL. Cere-
sonography. Posterior fossa anatomy is un- noses of possible inferior vermian agenesis brovascular accidents in neonates treated with
equivocally better displayed using the postero- (Fig. 5) that deserve mention. These errors extracorporeal membrane oxygenation: sonographic–
lateral fontanelle, as are posterior fossa were made as a result of misinterpretation of pathologic correlation. AJR 1989;153:355–361
566 AJR:174, February 2000

7. Sonography of Neonatal Posterior Fossa Abnormalities
2. Taylor GA, Fiktz CR, Glass P, Short, BL. CT of sound: anatomic and sonographic correlation. 12. Grunnet ML, Shields WD. Cerebellar hemor-
cerebrovascular injury after neonatal extracorporeal Early Hum Dev 1995;42:141–152 rhage in the premature infant. J Pediatr 1976;88:
membrane oxygenation: implications for neurode- 8. Buckley KM, Taylor GA, Estroff JA, Barnewolt 605–608
velopmental outcome. AJR 1989;153:121–126 CE, Share JC, Paltiel HJ. Use of the mastoid fon- 13. Martin R, Roessmann U, Fanaroff A. Massive in-
3. Babcock DS, Han BK, Weiss RG, Ryckman RC. tanelle for improved sonographic visualization of tracerebellar hemorrhage in low-birth-weight in-
Brain abnormalities in infants on extracorporeal the neonatal midbrain and posterior fossa. AJR fants. J Pediatr 1976;89:290–293
membrane oxygenation: sonographic and CT 1997;168:1021–1025 14. Merrill J, Piecuch RF, Fell SC, Barkovich AJ, Gold-
findings. AJR 1989;153:571–576 9. Shuman WP, Rogers JV, Mack LA, Alvord EC, stein RB. A new pattern of cerebellar hemorrhages
4. Sudakoff G, Montazemi M, Rifkin M. The fora- Christie DP. Real-time sonographic sector scan- in preterm infants. Pediatrics 1998;102:62–66
men magnum: the underutilized acoustic window ning of the neonatal cranium: technique and nor- 15. Laing FC, Frates MC, Brown DL, Benson CB, Di
to the posterior fossa. J Ultrasound Med 1993;4: mal anatomy. AJR 1981;137:821–828 Salvo DN, Doubilet PM. Sonography of the fetal
205–210 10. Grant EG, Borts FT, Schellinger D, McCullough posterior fossa: false appearance of mega-cisterna
5. Maertens P. Imaging through the posterior fon- DC, Sivasubramanian KN, Smith Y. Real-time ul- magna and Dandy-Walker variant. Radiology
tanelle. J Child Neurol 1989;4 [suppl]:S62–S67 trasonography of neonatal intraventricular 1994;192:247–251
6. Anderson N, Fulton J. Sonography through the hemorrhage and comparison with computed to- 16. Barkovich AJ, Kjos BO, Norman D, Edwards
posterior fontanelle in diagnosing neonatal intra- mography. Radiology 1981;139:687–691 MS. Revised classification of posterior fossa
ventricular hemorrhage. AJNR 1991;12:368–370 11. Yousefzadeh D, Naidich T. US anatomy of the cysts and cystlike malformations based on the re-
7. Anderson NG, Hay R, Hutchings M, Whitehead posterior fossa in children: correlation with brain sults of multiplanar MR imaging. AJR 1989;153:
M, Darlow B. Posterior fontanelle cranial ultra- sections. Radiology 1985;156:353–361 1289–1300
AJR:174, February 2000 567