FETUS, PLACENTA,

AND NEWBORN

Orbital diameters: A new parameter for prenatal

diagnosis and dating

KARA L. MAYDEN, R.T., R.D.M.S.

MARGE TORTORA, R.D.M.S.
RICHARD L. BERKOWITZ, M.D.
MICHAEL BRACKEN, PH.D.
JOHN C. HOBBINS, M.D.
New Haven, Connecticut

Fetal inner orbital diameter and outer orbital diameter were measured in the occipital transverse and
occipital posterior positions in 180 normal pregnancies. The outer orbital diameter was found to be
closely related to the biparietal diameter. After a nomogram had been constructed, 463 patients
whose fetuses were at risk for various anomalies were studied. In two cases, hypotelorism was
suspected because the orbital measurements fell below the 95% confidence limits, and this
diagnosis was confirmed at the time of delivery. The conclusion is that, in situations in which fetal
head position precludes accurate biparietal diameter determinations, the outer orbital diameter can
often be obtained and used to estimate that variable and, therefore, gestational age. Furthermore,
the nomogram constructed has proved to be a useful tool in diagnosing hypotelorism in utero.
Theoretically, this should also be true of fetal hypertelorism. (AM. J. OBSTET. GYNECOL. 144:289,
1982.)

WITH SUBSTANTIALLY improved ultrasound imag- the biparietal diameter, is virtually impossible to obtain
ery it is now possible to identify previously inaccessible when the fetal head is facing straight up or down. The
structures, such as fetal orbits. The ability to study orbits, however, can be identified and measured in an
these structures should permit the diagnosis of condi- occiput posterior position, and it is possible, therefore,
tions associated with hypotelorism (e.g., holoprosen- that orbital diameters could be used to date pregnan-
cephaly) and hypertelorism (e.g., the “fetal hydantoin cies in lieu of biparietal diameter measurements.
syndrome”) which affect the inner and outer orbital Since the only standard available today for normal
diameters.’ bony inner orbital diameter and outer orbital diameter
The standard measurement for dating pregnancies, was established for term infants through childhood,’
the following study was undertaken to establish an
From the Departments of Obstetrics and Gynecology, in vivo nomogram comparing the orbital diameters to
Epiakmiology, and Public Health, Yale University School the biparietal diameter. An attempt was also made to
of Medicine.
identify, and follow-up at delivery, fetuses whose or-
Received for publication February 8, 1982.
bital diameters were outside the normal range for these
Revised May 10, 1982.
parameters.
Accepted May 27, 1982.
Reprint requests: Kara L. Mayden, Department of Methods and material
Obstetrics and Gynecology, Yale University School of
Medicine, 333 Cedar St., P. 0. Box 3333, New Haven, The in vivo inner and outer orbital diameters were
Connecticut 06510. measured in 180 normal singleton fetuses at various

0002-9378/82/190289+09$00.90/0 0 1982TheC.V.Mosby Co. 289

290 Mayden et al. October 1. 1982
Am. J. Obstet. Gynecol.

With the head in the occipitotransverse position, the

transducer can be placed in two possible planes: (1)
along the coronal plane, approximately 2 cm posterior
to the glabella-alveolar line (Figs. 1A and 1 B), or (2)
along the orbitomeatal line, approximately 2 to 3 cm
below the level of the biparietal diameter (Figs. 2A and
2B). In both of these views, the midline, orbital rings,
nasal processes, and portions of the maxillae can be
demonstrated.
Both the inner orbital diameter and the outer orbital
diameter were measured from the leading edge to fall-
ing edge (outer to outer) by means of electronic cali-
pers. The outer orbital diameter was measured from
the lateral border of the orbit to the opposite lateral
border. The inner orbital diameter was measured from
the medial border of the orbit to the opposite medial
border (Figs. lB, 2B, and 3B).
With the head in the occipitoposterior position (3), a
biparietal diameter cannot be obtained. However, the
orbits can be identified and measured (Figs. 3A and
3B). The transducer was placed in a plane that tran-
sected the occiput, orbits, and nasal processes. Mea-
TRANSDUCER PLANE A \ surements were obtained only when the fetal face was
directly perpendicular to the uterine wall, since mea-
Fig. 1A. Frontal view demonstrating a fetus in a vertex pre- surements in an oblique plane were considered to be
sentation with the fetal cranium in an occipitotransverse (No. unreliable.
1) Position. The transducer is placed along the coronal plane
(approximately 2 cm posterior to the glabella-alveolar line). In order to assess the reliability of both orbital di-
ameter measurements, six abortuses between 16 and 22
weeks’ gestation were studied by ultrasound. Each
times during pregnancy. Gestational age was deter- abortus was suspended in a water bath and positioned
mined by the biparietal diameter or by the last men- to provide occipitotransverse, occipitoposterior, and
strual period when the biparietal was unobtainable be- occipitoanterior views. Between four and nine mea-
cause of an occipitoposterior position. The biparietal surements were obtained to compare the orbital mea-
diameter ranged from 22 mm (12.6 weeks’ gestation) to surements in each of three views and to see whether the
89 mm (36.4 weeks, according to the Yale nomogram).3 orbital diameters could be consistently and accurately
In addition, orbital diameters were studied in 463 pa- reproduced for each fetal head position. While mea-
tients referred for comprehensive ultrasound evalua- surements were being obtained, the resulting digital
tion between November, 1979, and May, 1981, because information was hidden from the “blinded” operator.
their fetuses were at risk for a variety of congenital
anomalies. Six patients who were specifically at risk for Results
conditions associated with hypotelorism or hypertelor- Several statistical models were fitted to the data in
ism were also evaluated. order to describe the growth of the orbital diameters in
Equipment used in the study included a digital terms of increasing biparietal diameter. A curvilinear
B-scanner (Picker, North Haven, Connecticut), and model, reflecting more rapid growth in the earlier ges-
any of three linear-array real-time systems (ADR, tational ages and somewhat slower growth later, pro-
Tempe, Arizona; General Electric, Milwaukee, Wis- vided the best fit to both the outer orbital diameter and
consin; Picker, North Haven, Connecticut). inner orbital diameter data (Table I). The observed
The fetal orbits were identified in three different data points, predicted average growth curves, and 95%
fetal head positions: occipitotransverse (n = 158); oc- confidence limits for both orbital diameters are plotted
cipitoposterior (n = 19); and occipitoanterior (n = 3). against biparietal diameter in Fig. 4.
Because of small numbers, and the inability to accu- Our analysis reveals that the outer orbital diameter is
rately define the orbital margins, the occipitoanterior much more strongly related to biparietal diameter than
position was abandoned for this analysis. is the inner orbital diameter (shown by the R* statistic in
Volume 144 Orbital diameters 291
Number 3

Fig. 1B. Sonogram through the fetal cranium in an occipitotransverse (No. 1) position. Nasal pro-
cesses (n), maxillae (m), fetal thorax (r;T), anterior uterine wall (AUW). The inner orbital diameter
(loo) is measured from the medial border of the orbit to the opposite medial border. The outer
orbital diameter (double white arrow~J is measured from the lateral border of the orbit to the opposite
lateral border.

Table I. Best-fitting regression model for orbital diameters

Dependent
W)
variable Independent
(Jo
variable Intercept
(1)
Biparietal
(a)
diameter Biparietul
(6)
diameter*
l- R *
Inner orbital diameter Biparietal diameter -2.961 0.455 -0.002 0.761
(1.589) (0.061) (0.001)
Outer orbital diameter Biparietal diameter -4.128 0.978 - 0.003 0.927
(2.406) (0.092) (0.001)

All measurements are in millimeters. The figures in parentheses are standard deviations.
*The regression equation is Y =I +ax + bx2.

Table I). Moreover, the outer orbital diameter is very tion of the outer orbital diameter over the inner orbital
highly correlated with biparietal diameter (R* = 0.927). diameter, since the length of the latter is observed to
By means of the above equation, it is possible to tabu- change only minimally, especially after the twenty-
late the inner orbital diameter and outer orbital diame- eighth gestational week. From Table II it is possible to
ter measurements with their corresponding biparietal estimate biparietal diameter and, therefore, gestational
diameters and estimated ;estational ages as defined by age if the orbits but not the biparietal diameter can be
the Yale nomogram.3 This is presented in Table II. visualized by ultrasound.
This table also demonstrates the superior discrimina- The three fetal head positions were analyzed sepa-
292 Mayden et al. October 1, 1982
Am. J. Obstet. Gynecol.

Table II-Cont’d
Biparietul diameter Weeks’ Inner orbital Outer orbital
(cm) gestution diameter (cm) diameter (cm)

4.4 19.4 1.3 3.2

4.5 19.9 1.3 3.3
4.6 20.4 1.3 3.4
4.7 20.4 1.3 3.4
4.8 20.9 1.4 3.5
4.9 21.3 1.4 3.6
5.0 21.3 1.4 3.6
5.1 21.8 1.4 3.7
5.2 22.3 1.4 3.8
5.3 22.3 1.5 3.8
5.4 22.8 1.5 3.9
5.5 23.3 1.5 4.0
5.6 23.3 1.5 4.0
5.7 23.8 1.5 4.1
5.8 24.3 1.6 4.1
5.9 24.3 1.6 4.2
6.0 24.7 1.6 4.3
6.1 25.2 1.6 4.3
6.2 25.2 1.6 4.4
6.3 25.7 1.7 4.4
6.4 26.2 1.7 4.5
6.5 26.2 1.7 4.5
6.6 26.7 1.7 4.6
6.7 27.2 1.7 4.6
6.8 27.6 1.7 4.7
’ TRANSDUCER PLANE -/ 6.9 28.1 1.7 4.7
7.0 28.6 1.8 4.8
7.1 29.1 1.8 4.8
Fig. 2A. Frontal view demonstrating the fetal cranium in an 7.3 29.6 1.8 4.9
occipitotransverse (No. 2) position. The transducer is placed 7.4 30.0 1.8 .5.0
along the orbitomeataf line (approximately 2 to 3 cm below 7.5 30.6 1.8 5.0
the level of the biparietaf diameter). 7.6 31.0 1.8 5.1
7.7 31.5 1.8 5.1
7.8 32.0 1.8 5.2
Table II. Predicted biparietal diameter 7.9 32.5 1.9 5.2
and weeks’ gestation from the inner 8.0 33.0 1.9 5.3
and outer orbital diameters 8.2 33.5 1.9 5.4
8.3 34.0 1.9 5.4
Bipatietnl diameter Weeks’ Inner orbital Outer orbital 8.4 34.4 1.9 5.4
(ml gestation diameter (cm) diameter (cm) 8.5 35.0 1.9 5.5
8.6 35.4 1.9 5.5
1.9 11.6 0.5 1.3 8.8 35.9 1.9 5.6
2.0 11.6 0.5 1.4 8.9 36.4 1.9 5.6
2.1 12.1 0.6 1.5 9.0 36.9 1.9 5.7
2.2 12.6 0.6 1.6 9.1 37.3 1.9 5.7
2.3 12.6 0.6 1.7 9.2 37.8 1.9 5.8
2.4 13.1 0.7 1.7 9.3 38.3 1.9 5.8
2.5 13.6 0.7 1.8 9.4 38.8 1.9 5.8
2.6 13.6 0.7 1.9 9.6 39.3 1.9 5.9
2.7 14.1 0.8 2.0 9.7 39.8 1.9 5.9
2.8 14.6 0.8 2.1
2.9 14.6 0.8 2.1
3.0 15.0 0.9 2.2
3.1 15.5 0.9 2.3 rately by mean,s of the simple linear growth model
3.2 15.5 0.9 2.4 (Y = I + ax) for this comparison (Table III). For each
3.3 16.0 1.0 2.5
position, the outer orbital diameter was a better predic-
3.4 16.5 1.0 2.5
3.5 16.5 1.0 tor of biparietal diameter than was the inner orbital
3.6 17.0 1.0 z:; diameter. For the outer orbit, the occipitotransverse
3.7 17.5 1.1 2.7
3.8 17.9 1.1 2.8 (No. 2) position showed some additional advantage
4.0 18.4 1.2 3.0 over the other two positions, but for all three the linear
4.2 18.9 1.2 3.1 model for the outer orbital diameter provided a very
4.3 19.4 1.2 3.2
good fit for the data.
Volume 144
Number 3 Orbital diameters 293

Fig. 2B. Smogram through the fetal cranium in an occipitotransverse (No. 2) position. The midline
(ML), nasal processes (n], outer orbital diameter (OOD, duubb small w-rows), and inner orbital diameter
(ZOD, white bars) can be demonstrated.

Table III. Regressional model for orbital diameters for three fetal orbit positions

Dependent variable Biparietal diameter

(Y) la) R ‘*

Occipitotransverse (I) Inner orbital diameter -0.832 0.306 0.726

Outer orbital diameter 1.789 0.695 0.850
Occipitotransverse (II) Inner orbital diameter 4.039 0.195 0.724
Outer orbital diameter 5.504 0.606 0.922
Occipitoposterior (III) Inner orbital diameter 4.241 0.167 0.704
Outer orbital diameter 8.107 0.557 0.853

*The regression equation is Y = I + ax.

294 Mayden et al. October 1, 1982
Am. J. Obstet. Gynecol.

ANTERIORUTERINEWALL
/-

l- POSTERIORUTERINEWALL
Fig. SA. Side view demonstrating the fetus in an occipitoposterior position. The transducer is placed
in a plane that transects the occiput, orbits, and nasal processes.

Fig. 3B. Monogram through the fetal cranium in an occipitoposterior (No. 3) position. Placenta(P),
occiput @urge arrou), nasal processes (small uhite avows), outer orbital diameter (x’s), and inner
orbital diameter (circular marks). Note the electronic digital calculation for the inner orbital diameter.
Volume 144 Orbital diameters 295
Number 3

65r .

60
t

/-
4-0’ Y = -2.961+ 0.455x - 0.002x’
01”““““““““’
22 30 36 46 54 62 70 78 66 94
BIPARIETAL DIAMETER (mm)
Fig. 4.Nomogram comparing the inner orbital diameter and outer orbital diameter to the biparietal
diameter ( 180 individual measurements).

The minimum number of in vitro observations for surements for femur length4 and abdominal circum-
any one position on a single abortus was four. The ferences). Ventricular dilatation was also noted. The
coefficient of variation for the inner orbital diameter outer orbital diameter was even further discrepant,
ranged from 9.3% to 12.9%, and for the outer orbital being 7 weeks less than the biparietal diameter. The
diameter, from 2.1% to 8.8%. Thus, there appears to inner orbital diameter was compatible with a mean ges-
be relatively little variation between several observa- tational age of 13.6 weeks’ gestation. Vaginal delivery
tions of a single specimen when viewed from the same occurred at term. The head circumference measured
position. 27 cm, which was compatible with severe microcephaly.
Of the 463 patients studied who were at risk for The inner orbital diameter and outer orbital diameter
congenital anomalies, abnormal orbital diameters were measured 1.5 and 4.5 cm, respectively, and confirmed
detected in three fetuses. In each of these cases, the the diagnosis of hypotelorism. The infant was diag-
orbital diameters were below the fifth percentile, nosed as having cebocephaly (microcephaly, hypotelor-
and hypotelorism was suspected (Fig. 5). The first ism, a single, centrally placed ventricular cavity, and a
pregnancy was terminated and the abortus was not ex- tubular nose without a septum) (Fig. 6).
amined. The second pregnancy was also electively Ultrasound was employed to measure the orbital di-
terminated. The abortus was a male and had multiple ameters in six fetuses at risk for hypotelorism (cyclopia,
congenital anomalies, including unilateral renal one; holoprosencephaly, two) and hypertelorism (“fetal
agenesis, bowed legs, an intraventricular septal defect, hydantoin syndrome,” one; Apert’s syndrome, one;
imperforate anus, and a prominent nose. Hypotelor- and G syndrome-hypospadias-dysphagia or Opitz
ism was clinically evident, although, regrettably, the Frias syndrome, one). Ultrasound measurements of
orbital distances were not measured. In the third case, both the inner orbital diameter and outer orbital di-
the fetus was suspected to have microcephaly at 33 ameter revealed normal diameters in each case, and at
weeks’ gestation because of a 6%-week discrepancy be- birth all orbital diameters were demonstrated to be
tween biparietal diameter and dates (validated by mea- within normal limits.
 296 Mayden et al. October 1, 1982
Am. J. Obstet. Gynecol.

Fig. 5. In the occipitotransverse (No. 2) position, hypotelorism is demonstrated. Outer orbital diame-
ter (OOD, cross bars); inner orbital diameter (IOD, small whit? arrows); occiput (large cunvd arrow).

Comment the diagnosis of fetal conditions in which hypotelorism

An excellent correlation was found to exist between or hypertelorism is a feature. In this study, three
outer orbital diameter and biparietal diameter in each fetuses were found to have outer orbital diameter and
of the three fetal head positions studied. The strongest inner orbital diameter measurements that were below
correlation was noted when the fetus was in the occipi- the fifth percentile. Although one abortus was lost to
totransverse position, but the results were also very ac- follow-up, hypotelorism was confirmed in two fetuses
ceptable when the head was facing - up. This close rela- at termination or delivery. In six fetuses at risk for
tionship between outer orbital diameter and biparietal ocular abnormalities, normal antenatal measurements
diameter may allow the physician to date pregnancies correctly predicted normal neonatal distances in all
when the latter measurement is impossible to obtain. cases.
Furthermore, by extrapolating biparietal diameter val- As is true of all measurements made with ultrasound,
ues from outer orbital diameter measurements, one outer orbital diameter and inner orbital diameter de-
can estimate fetal weight from existing formulas by terminations must be performed precisely or the data
utilizing biparietal diameter as a variable.;’ will be misleading. For example, tangential cuts through
Orbital diameter measurements are a useful tool in the orbits can easily produce erroneous measurements.
Volume 144 Orbital diameters 297
Number 3

Fig. 6. Neonate with cebocephaly.

Also, it is occasionally difficult to, define accurately the fetal orbital architecture in patients at risk for ocular
distal orbital margin in the occipital transverse position abnormalities.
because of acoustic shadowing from the nose. Never-
theless, with these caveats in mind, we have found this We are grateful to Dr. Maurice J. Mahoney for his
technique to be valuable in evaluating gestational age in aid and advice and Ms. Elizabeth Kiefer for the techni-
lieu of biparietal diameter, and essential in assessing cal illustrations.

REFERENCES
stetrics and Gynecology, Baltimore, 1977, The Williams &
1. Bergsma, D., editor: Birth Defects Compendium, ed. 2, Wilkins Co.
New York, 1980, Alan R. Liss, Inc., pp. 526527, 433- 4. Hobbins, J. C., Bracken, M. B., and Mahoney, M. J.: Diag-
434. nosis of fetal skeletal dysplasias with ultrasound, AM. J.
2. Smith, D. W.: Recognizable patterns of human malforma- OBSTET. GYNECOL. 142:306, 1982.
tion, in Schaeffer, A. J., editor: Major Problems in Clinical 5. Shepard, M. J., Richards, V. M., Berkowitz, R. L., Warsof,
Pediatrics, ed. 2, Philadelphia, 1976, W. B. Saunders, pp. S., and Hobbins, J. C.: An evaluation of two equations for
467-468. predicting fetal weight by ultrasound, AM. J. OBSTET. GY-
3. Hobbins, J, C., and Winsberg, F.: Ultrasonography in Ob- NECOL. 142:47. 1982.