Normative Measurements of

Orbital Structures Using CT

AIiOzgen’ OBJECTIVE. The purposes of this study were to establish criteria for the diameters of nor-
mal extraocular muscles, to determine the normal position of the globe as revealed by CT. and to
Macit Ariyurek
investigate the effects of age and sex on these structures.
SUBJECTS AND METHODS. Diameters of extraocular muscles, distance from the in-
terzygomatic line to the posterior margin of the globe. width of the optic nerve-sheath com-
plex. and length of the interzygomatic line were calculated for 2(X) normal orbits of 100
American Journal of Roentgenology 1998.170:1093-1096.

patients on axial and direct coronal CT images. Effects of age and sex on muscle diameters and
globe position were analyzed.
RESULTS. Normal ranges for the diameters (mean ± 2SDs) of extraocular muscles were
medial rectus. 3.3-5.0 mm: lateral rectus, I .7-4.8 mm: inferior rectus. 3.2-6.5 mm; and superior
group. 3.2-6. 1 mm. The normal position of the globe was 9.4 mm behind the interzygomatic
line (range. 5.9-l2. mm). The mean diameters ofthe extraocular muscles and the length of the
interzygomatic line in male patients were significantly larger than in female patients (p < .()0l).
Statistically significant correlation was found between age and the diameters of the inferior and
lateral rectus muscles(r= .32.p = .013: and r= .23.p = .048. respectively).
CONCLUSION. Our results may be important in interpreting CT scans of the orbit be-
cause. tO our knowledge. no reliable normative data exist regarding these orbital structures.

C T has
nique
become
for
a widely
imaging of
used
the
tech-
orbit
were
ditions
clinical suspicion
of the paranasal
of tumor
sinuses
and tumorlike
fr which
con-
surgical

since the l970s. Although normal treatment was contemplated. All patients in this
study were free ofclinical evidence or history of en-
anatomy of the orbital structures as seen on
docrine disease or any orbital disorder. Asymmetric
CT is described in many references. in a re-
scans. scans with artifacts for any reason (dental
view of the literature we could not find reli-
material. eye motion. etc. ) that may cause errors in
able quantitative normative CT data for orbit measurements. and CT scans with abnormal
extraocular muscle diameters, which may be orbital findings were excluded. Two hundred nor-
important in evaluating an individual with mal orbits of 1(X) patients (44 men and 56 women)
suspected orbital disease such as Graves’ orb- were evaluated in this study. The age range of the
itopathy. In this situation. the ranges of diame- patients was 18-70 years (mean. 4 I years).
tens of healthy extraocular muscles may be
critical to accurately assess the muscles as Methods

normal. The same information is needed to di- Both axial and direct coronal, 3-mm-thick non-

agnose whether exophthalmos is present. In overlapping contiguous sections were obtained in

this study. we present the diameters of healthy

all patients with a CT scanner (Tomoscan SR 7(X)O;
Philips. Eindhoven. the Netherlands). Axial scans
Received July 22, 1997; accepted
September 29, 1997.
after revision
extraocular
protuberance
muscles
of the
and
globe
the normal
from the
anterior
orbit as
were obtained at an angle of - 10#{176}
to -I 5#{176}
relative to
the orbitomeatal plane. Direct coronal scans were
1 Both authors: Department of Radiology, Hacettepe seen on CT.
obtained almost perpendicular to the axial plane.
University School of Medicine, Sihhiye 06100, Ankara, Patients were asked to maintain forward gaze and
Turkey. Address correspondence to A. Ozgen, Sporcular
Sitesi, llkyerlesim Mahallesi, 438. sokak, No. 12, Batikent gentle eye closure during the scans to prevent asym-
Subjects and Methods
06370, Ankara, Turkey. metric extraocular muscle contraction. Axial and
Patients
coronal scans of the orbits were postprocessed with
AJR 1998;170:1093-1096
All CT images in this study were obtained from a 5 1 2 x 5 12 matrix correspnding to a field of view
0361-803X198/1704-1093 patients who were referred to our department for of I 20 x I 20 mm. Because every change in the win-
© American Roentgen Ray Society CT of the paranasal sinuses. Indications ton CT dow settings resulted in different measurements in

AJR:17O, April 1998 1093

 Ozgen and Ariyurek

the same CT scan, all scanning was perkwmed at

constant window level and width settings of 50 and
250 H. respectively. To determine the normal globe
position in CT. the interzygomatic line at the midg-
lobe section was used as a reference line. The length
ofthe interzygomatic line and the perpendicular dis-
tance between the interzygomatic line and the pos-
tenor margin of the globe at the midglobe section
were measured on axial scans. The width of the op-
tic nerve-sheath complex was also measured per-
pendicular to its course in the axial Cl’ sections
where the middle portion of the nerve was visual-
ized (Fig. I). Because they could not be reliably dis-
tinguished from each other, the superior rectus and
the levator palpebrae superior muscles were mea-
sured together as a single muscle group-the supe-
nor muscle group-before they become separate
with a fat plane between them. Vertical diameters of
the superior muscle group and the inferior rectus
muscle were measured on coronal scans (Fig. 2).
Horizontal diameters of the lateral and medial rec- Fig. 1.-Axial CT scan of 32-year-old woman at midglobe level shows interzygomatic line (A-B) and perpendicular
tus muscles were measured on both scan planes. distance from interzygomatic line to posterior margin of globe (C-D). Black lines indicate maximum diameters of me-
dial rectus muscle, lateral nectus muscle, and optic nerve-sheath complex, where measurements were made.
Measurements were done directly on the magnified
American Journal of Roentgenology 1998.170:1093-1096.

hard-copy images with the same magnification fac-

ton for all patients and then were converted to true relations using Pearso&s correlation. Ninety-five mean values and normal ranges for muscle di-
size in millimeters. In this study. the diameter of percent normal cutoff values were computed by ameters and globe position were obtained us-
each muscle was measured at its maximum. adding and subtracting two SDs from the mean. ing only right orbit data.
All statistics in this study were done using SPSS The mean diameters of extraocular muscles
Statistics
for Windows (SPSS. Chicago, IL).
Diameters of the medial and lateral rectus mus- and the mean length of the interzygomatic line
des on axial scans and diameters of the inferior in male patients were significantly larger than
rectus muscle and the superior muscle groups Results those in female patients (Table 2). whereas the
were summed to achieve a total of all muscles in Mean values and normal ranges for the di- ratio of the length of the interzygomatic line to
any individual. The ratio of the length of the in- ameters of extraocular muscles, the distance the diameter of each muscle had no statisti-
terzygomatic line to the diameter of each muscle between male and
between the interzygomatic line and the pos- cally significant difference
was calculated for all patients. The ratio of diame-
terior margin of the globe. the width ofthe op- female patients (p > .06). The length ofthe in-
tens measured on coronal images to diameters
tic nerve-sheath complex, the length of the terzygomatic line was positively correlated
measured on axial images was calculated for the
interzygomatic line, and the difference in with the diameter of each muscle (p .26,
medial and lateral rectus muscles. The paired-
ranges between right and left orbit data are r .009). Neither the width of the optic nerve-
samples t test was used to compare data obtained
from the right and left orbits. The independent- given in Table I. No statistically significant sheath complex nor the globe position showed
samples t test was used to compare data obtained difference was found between data for the a statistically significant difference between
from male and female patients. We calculated cor- right and left orbits (p < .001 ). Thereafter. male and female patients.

Fig. 2.-Sequential coronal CT scans of 23-year-old woman with sinusitis. Black lines indicate maximum diameters of muscles, where measurements were made.
A-C, Scans show measurement of superior muscle group (A), medial rectus muscle (B), and inferior and lateral rectus muscles (C).

1094 AJR:170, April1998

 CT of Orbital Structures

normal criteria. We think that the limited

#{149}V1UNormal Orbital Measurements as Seen on CT
number of control patients in the study by Nu-
Mean Normal Range (±2 SOs) Difference Rangea gent et al. raises questions about the reliability
Measurement
(mm) (mm) (mm)
of the normal values. The researchers ob-
Muscle tamed coronal sections with variable angles to

Medial rectus 4.2 3.3-5.0 0.5 avoid dental amalgams or because of limita-
Lateral nectus 3.3 1.7-4.8 0.7 tions in neck mobility in both groups. The
window level and width settings of that study
Superior group 4.6 3.2-6.1 0.9
(70 and 350 H, respectively) were also differ-
Inferior rectus 4.8 3.2-6.5 1.1
ent from our settings. Data for the lateral nec-
Sum of all muscles 16.9 13.1-20.7 1.4
tus muscle. the superior muscle group. and the
Globe positionb 9.4 5.9-12.8 1.0
sum of all muscles were significantly smaller
Optic nerve-sheath complex 4.4 3.2-5.6 0.7
in that study than in ours.
lnterzygomatic line 99 91-108 Some volumetric studies in patients with
Note-Medial and lateral rectus muscles were measured on axial plane, and superior group and inferior rectus muscle were Graves’ disease were also published. but the
measured on coronal plane at window level and width settings of 50 and 250 H, respectively. The diameter of each muscle was
volumetric method of quantitative evaluation
measured at its maximum.
aDifference range between right and left orbit data in 95th percentile limits. of the extraocular muscle volumes is time-
bperpendicular distance between interzygomatic line and posterior margin ofthe globe. consuming and needs additional specific hard-
ware and software I 10. 1 1 J. These volumetric
We found no consistent correlation between muscles are imaged for these reasons on CT studies also lack a sufficient number of control
age and the width of the optic nerve-sheath and evaluated qualitatively and subjectively
American Journal of Roentgenology 1998.170:1093-1096.

in subjects to establish normative data on cx-

complex or the globe position. Statistically respect to their sizes in addition to their shapes traocular muscle volumes.
significant correlation was found between age and borders I 1-7]. In the literature, quantita- We think that measuring one diameter of an
and the diameter of the inferior and the lateral tive mean values for the diameters of healthy extraocular muscle is the only
way to quanti-
rectus muscles (r = .32, p = .013: and r = .23, extraocular muscles were established in two tatively evaluate the size of that muscle in a
p = .()48. respectively). That is. the inferior studies. In 1980. Jacobs et al. 8] published or- practical and easy-to-use method. Measuring
and lateral rectus muscles tend to be slightly bit measurements derived from 2(X) normal diameters of extnaocular muscles in the axis
larger in older patients. The ratio of the diame- orbits imaged with 3- to 8-mm-thick CT sec- perpendicular to the orbital wall is also used
ten of the lateral rectus muscle measured on tions and relatively low resolution in which in CT imaging worldwide 19. 12-14]. We pre-
the coronal plane to the diameter measured on the researchers could image the extraocular ferred to measure only one cross-sectional di-
the axial plane was I .53 (range. I .03-2.46). muscles with an incidence varying from 38% ameter of the muscle-the horizontal diameter
The same ratio tbr the medial rectus muscle to 98%. The normative data published in that for the lateral and medial rectus muscles and
was 1 . 16 (range, 0.95-1.40). book show significant differences from our re- the vertical diameter for the superior muscle
suIts for the diameters of extnaocular muscles. group and the inferior rectus muscle.
Only the width of the optic nerve-sheath com- As a rule. visualization of an eye muscle is
Discussion plex has the same mean value. We believe the optimal ifthe plane ofthe section is parallel to
Enlargement of the extraocular muscles normative data published in that book have the course ofthe muscle I l5J. That means that
may be due to primary neoplasm. nonspecific lost their validity because of major technical imaging of the medial and lateral rectus mus-
inflammation. metastatic tumor, vascular mal- improvements in CT imaging. Another study. des is best performed on axial CT sections.
formation. infection. acromegaly. and trauma. by Nugent et al. [91, involved Graves’ orbitop- Therefore. we prefer to use data obtained on
as well as to Gnaves orbitopathy. the most athy and included 20 patients with normal CT axial sections to calculate the normal values
common cause I 1-31. In general. extraocular imaging findings of the orbit for establishing for these muscles. The wide range of the ratio
of diameters measured on coronal sections to
lameters of Healthy Extraocular Muscles the Length of the
and those measured on axial sections confirmed
terz omatic Une In Male and Female Patients as Seen on CT that imaging of these muscles would not be
satisfactory if only coronal sections were ob-
tamed. We measured diameters of the inferior
and superior group muscles on the coronal
plane. Direct sagittal sections could result in
better images of these muscles. but sagittal ne-
constructions from our data would not have
enough resolution to achieve accurate mea-
surements. Besides, axial and coronal planes
are routinely used in CT imaging of the orbit,
whereas the sagittal plane is not.
Our study revealed that the mean diameters
Note-Me teral rectus muscles were measured on axial plane, and superior group and inferior rectus muscle were
of the extnaocular muscles in male patients are
measured on coronal plane at window level and width settings of 50 and 250 H, respectively. The diameter of each muscle was
measured at its maximum. significantly larger than those in female pa-
aMean values for male and female patients are statistically different lp .0071. tients. The length of the interzygomatic line is

AJR:170, April 1998 1095

 Ozgen and Ariyurek

positively correlated with the diameter of each would be valid for specific window level and dardized A-scan echography and computerized to-
muscle, and the mean length of the interzygo- width settings, 50 and 250 H, respectively. mography scan measurements. Ophthalmology
1985:92:1351-1355
matic line in male patients is also significantly In conclusion, we present normal orbital
7. Barrett L. Glatt Hi. Burde RM. Gado MH. Optic
more than in female patients, whereas the ra- measurements that may help observers to ac-
nerve dysfunction in thyroid eye disease. Radio!-
tio of the length of the interzygomatic line to curately assess enlargement or atrophy of ex- og\, 1988:167:503-507
the diameter of each muscle does not show a traocular muscles and to determine whether 8. Jacobs L. Weisberg LA, Kinkel WR. Computerized
statistically significant difference between exophthalmos (or enophthalmos) is present in tomography (1 the orbit and se!!a turn’ica. New
male and female patients. We think that the a practical quantitative method. Because the York: Raven. 1980:27-85
9. Nugent RA. Belkin RI, Neigel JM, et al. Graves
difference between the diameters of extraocu- normative data for any sex (Table 2) show
orbitopathy: correlation of CT and clinical find-
lar muscles in men and women depends on equal to or less than 8% difference from the
ings. Radiology 1990;177:675-682
the head size (which may be represented by normative data for the population (Table I),
10. Forbes G, Gorman CA, Gehring D. Baker HL.
the length of the interzygomatic line) rather the latter may be used as the normative data Computer analysis of orbit fat and muscle volumes
than the sex of the patient. for all patients in a more practical method, al- in Graves ophthalmopathy. AJNR 1983:4:737-740
The sum of the diameters of all extraocu- though normal values for each sex may pro- 1 1. Forbes G. Gorman CA. Brennan MD, Gehring DO,
lar muscles may help as an overall index to vide slightly more accurate determination. Ilstnip DM, Earnest F. Ophthalmopathy of Graves’
disease: computerized volume measurements of the
evaluate patients’ measurements, particularly
orbit fat and muscle. AJNR 1986;7:65l-656
in patients with Graves’ disease, in which the References 12. Hallin ES. Feldon SE. Graves’ ophthalmopathy. 1.
diameter of an individual muscle remains in 1. Trokel SL, Hilal SK. Recognition and differential di- Simple CT estimates of extraocular muscle vol-
the normal range although the muscle is ac- agnosis ofenlagod extraocular muscles in computed ume. BrJ Ophthaltnol 1988:72:674-677
tually minimally enlarged. Values for the tomography.Am J Op/u/ia/mo! 197987:503-5l2 13. Chen YL, Chang TC. Huang KM. Tzeng SS. Kao
limits of the differences between diameters 2. Rothfus WE. Curtin HD. Extraocular muscle en- SCS. Relationship of eye movement to computed
American Journal of Roentgenology 1998.170:1093-1096.

largement: CT review. Radiology 198.3;l5l:677-681 tomographic findings in patients with Graves oph-
of healthy contralateral extraocular muscles
3. Patrinely JR. Osbom AG. Anderson RL. Whiting thalmopathy. Ac’ta Ophthalmo! 1994:72:472-477
may also be helpful.
AS. Computed tomographic features of nonthyroid 14. Wilson RG. Pope RM, Michell Mi, Cannon R,
Some researchers evaluated diameters of extraocular muscle enlargement. Ophthalmology McGregor AM. The use of real-time ultrasound in
extraocular muscles with varying window set- 1989;96:l038-l047 Graves’ ophthalmopathy: a comparison with corn-
tings [12, 13, 16]. Every change in the win- 4. Yoshikawa K, Higashide 1, Nahase Y, Inoue 1, In- puted tomography. BrJ Radio! 1989:62:705-709
dow level and width settings results in oue Y, Shiga H. Role of rectus muscle enlargement 15. Unsold R. Newton TH. De Groot I. CF evaluation
different values with respect to the muscle in clinical profile of dysthyroid ophthalmopathy. of extraocular muscles: anatomic-CT correlations.
Jpn J Ophthalmol 1991;35:175-l8l Aibrechi Von Graefes Arch KIm Ophthalmo! 1980;
size. That means that window settings should
5. Hudson HL, Levin L, Feldon SE. Graves exoph- 214: 155-180
be the same to accurately compare the muscle
thalmos unrelated to extraocular muscle enlarge- 16. Feldon SE. Levin L. Liu 5K. Graves’ ophthalmopa-
sizes both between different patients and be- ment. Ophthalmology 1991;98:1495-l499 thy: correlation of saccadic eye movements with age.
tween different CT examinations of the same 6. Holt JE. O’Connor PS, Douglas JP. Byrne B. presence ofoptic neuropathy. and extnaocular muscle
patient. For this reason, our normative data Extraocular muscle size comparison using stan- volume. An’h Ophtha!mo! 1990:l08:1568-l57l

1096 AJR:170, April 1998