Clinical and Experimental Ophthalmology 2009; 37: 30–53 doi: 10.1111/j.1442-9071.2008.01822.
Review
Imaging for neuro-ophthalmic and orbital disease –
a review
Andrew G Lee MD,1,2,3 Michael C Johnson MD FRCSC,1 Bruno A Policeni MD4 and
Wendy RK Smoker MD FACR3,4
Departments of 1Ophthalmology and Visual Sciences, 2Neurology, 3Neurosurgery and 4Radiology, University of Iowa Hospitals and
Clinics, Iowa City, Iowa, USA
ABSTRACT
A literature review was performed by content experts in
neuro-ophthalmology and neuroradiology using a sys-
tematic English-language Medline search (1994–2008)
limited to articles with relevance to neuro-ophthalmic
and orbital imaging. The information covered in this
review includes: (i) the basic mechanics, indications
and contraindications for cranial and orbital computed
tomography and magnetic resonance (MR) imaging;
(ii) the utility and indications for intravenous contrast,
(iii) the use of specific MR sequences; (iv) the tech-
niques and ophthalmic indications for computed
tomography/MR angiography and venography; and (v)
the techniques and indications for functional MR
imaging, positron emission tomography scanning and
single photon emission computed tomography.
Throughout the review accurate and timely communi-
cation with the neuroradiologist regarding the clinical
findings and suspected location of lesions is emphasized
so as to optimize the ordering and interpretation of
imaging studies for the ophthalmologist.
Key words: computed tomography, diagnostic imaging,
magnetic resonance angiography, magnetic resonance
imaging, visual pathway.
BACKGROUND
Ophthalmologists may be the first clinicians to
evaluate the patient with an orbital or intracranial
! Correspondence: Dr Andrew G Lee, Department of Ophthalmology, 200 Hawkins Drive, PFP, University of Iowa Hospitals and Clinics, Iowa City, IA
52242, USA.
Email: andrew-lee@uiowa.edu
Received 23 January 2008; accepted 26 June 2008.
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
structural lesion. Therefore, it is important for the
ordering ophthalmologist to understand the basic
mechanics, indications and contraindications for
specific imaging studies, such as computed to-
mography (CT) and magnetic resonance (MR)
imaging (MRI). Over the past decade, these
imaging techniques have evolved and continue to
improve. There are also many techniques that have
recently emerged and that are of interest to the
ophthalmologist.
This monograph provides a summary of the most
commonly used imaging techniques in ophthalmo-
logy. We update and summarize the recent literature
on the various modalities (e.g. MR angiography
[MRA], CT angiography [CTA], MR venography
[MRV] and CT venography [CTV]) and specific
sequences (e.g. fat suppression, fluid attenuation
inversion recovery (FLAIR), gradient recall echo
imaging [GRE], diffusion-weighted imaging [DWI],
perfusion-weighted imaging [PWI] and dynamic
perfusion CT [PCT]) used in neuro-ophthalmic and
orbital imaging. We also include a discussion of the
indications and contraindications for using contrast
in CT and MRI. Finally, we briefly review the oph-
thalmic indications for functional imaging (e.g. func-
tional MRI [fMRI], positron emission tomography
[PET] and single photon emission computed tomog-
raphy [SPECT]). Throughout this review, we empha-
size the critical importance of accurate and timely
communication of the clinical findings and pre-
sumed location of a lesion with the neuroradiologist
in order to acquire the best imaging study and
receive the best interpretation.
Neuro-ophthalmic and orbital imaging
METHODS
An English-language Medline search was performed
from 1994 to 2008 for articles on orbital or neuroim-
aging relevant to ophthalmology, including CT, MRI,
angiography (e.g. catheter, MRA and CTA), venog-
raphy (e.g. MRV and CTV), fMRI, PET and SPECT.
Two content experts from ophthalmology (AGL,
MCJ) selected relevant titles and abstracts and then
reviewed the pertinent papers with two neuroradi-
ology content experts (BAP, WRKS). Publications
prior to 1994 were included only if they added sig-
nificant or new information to the review. The infor-
mation from the review was collated into a clinical
summary of the available radiographic techniques of
interest to the ophthalmologist, with indications and
contraindications included for each study.
RESULTS
The review identified the following clinical sce-
narios for which an imaging study might be ordered
by an ophthalmologist: (i) unilateral or bilateral
visual loss (e.g. transient visual loss [amaurosis
fugax], unilateral or bilateral optic neuropathy, junc-
tional scotoma, bitemporal hemianopsia, homony-
mous hemianopsia or cortical blindness); (ii) efferent
pupillary defects (e.g. anisocoria due to Horner syn-
drome or third nerve palsy) or afferent pupillary
defects (e.g. relative afferent pupillary defect or light-
near dissociation of the pupils); (iii) proptosis (e.g.
thyroid eye disease, orbital tumours, idiopathic
orbital inflammation [i.e. orbital pseudotumour],
orbital cellulitis or carotid-cavernous fistula); (iv)
diplopia or external ophthalmoplegia; (v) lid abnor-
malities (e.g. lid retraction, lid lag, ptosis or orbital-
lid lesion); (vi) oscillopsia (e.g. nystagmus); (vii)
ophthalmoscopic abnormalities (e.g. papilloedema,
optic atrophy, optic nerve hypoplasia, optic disc head
drusen or choroidal folds); and (viii) ocular or orbital
trauma (e.g. intraocular/intraorbital foreign body or
suspected fracture). In this review we have elected to
exclude the neuroimaging evaluation of isolated
headache or facial pain as these are beyond the scope
of this work.
The three main imaging techniques for the brain
and orbit are conventional X-rays (e.g. skull film),
CT and MR scanning. Although the X-ray has been
known since Roentgen, the role of the traditional
skull or orbital X-ray has been supplanted in the
modern era by current CT and MR techniques. We
use the skull X-ray to prescreen patients with a
history or suspicion for metallic foreign body prior to
MR studies. As a general rule, patients who have an
indication for orbital imaging should undergo a CT
or an MR study, if not contraindicated, for improved
bone or soft-tissue resolution.
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
31
CT scanning
The CT scan, first introduced in 1972, is based on
standard X-ray attenuation by tissues of various
densities.1 Images are acquired by rapid rotation of
an X-ray tube around the patient. The transmitted
radiation is measured by a ring of detectors located
on the gantry. Data points obtained by the detectors
are analysed using a computer and the density, or
attenuation value, of the tissue at each point is
calculated. The image is reconstructed as a corre-
sponding matrix of picture elements (pixels), with
each pixel being assigned a numerical value. These
values are then compared with the attenuation value
of water and displayed on a scale of arbitrary units
called Hounsfield units. This scale assigns water as
an attenuation value of zero. Denser material (e.g.
lead) blocks, or attenuates, the X-ray beam and less
dense material allows the beam to pass through to
the detectors. By convention, the denser material
(e.g. bone) is ‘brighter’ (positive Hounsfield units)
on CT and less dense material (e.g. air) is ‘darker’
(negative Hounsfield units).
Early CT scanners acquired images one single
slice at a time. During the 1980s, technical advances
enabled the X-ray tube to rotate continuously in one
direction around the patient. This led to the intro-
duction of helical (or spiral) CT in 1989.2 Data could
now be continuously acquired while the patient was
simultaneously moved at a constant speed through
the gantry. The transmitted radiation took on the
form of a helix or spiral. This allowed larger ana-
tomical regions of the body to be imaged during a
single breath hold, less overall radiation exposure
and reduced movement artefacts. Faster scanning
also increased patient throughput. Multislice or
multidetector-row CT scanners are the current gen-
eration of scanners commercially available. They use
the principle of helical scanners, but incorporate
multiple rows of detector rings (from 4 to 64) and can
therefore acquire multiple slices per tube rotation
and increase the area covered in a given time by the
X-ray beam.3
The contrast material for CT scanning is iodinated
contrast. Prior allergic reaction to iodinated contrast
or a history of renal failure may be contraindications
to using contrast in CT.4–6 In general, the use of con-
trast improves the sensitivity and specificity of CT
scan interpretation.5 For most conditions in ophthal-
mology, contrast material should be included when
ordering a CT scan. However, there are a few excep-
tions to this recommendation. For example, a CT
scan that is being performed for acute intracranial or
intraorbital haemorrhage does not require contrast
material. In fact, the non-contrast study is superior
for assessing the hyperdensity of acute blood.7 In
addition, in cases of trauma (e.g. orbital, facial or
32
Figure 2. Idiopathic orbital inflammation. (a) Coronal and (b, c) axial post-contrast computed tomography images demonstrate enlarge-
ment of all extraocular muscles in the left orbit, best appreciated on the coronal image (a, white dots), whereas the axial images show
involvement of the tendinous insertion of the medial rectus muscle (b, short arrow) and a stranded appearance of the retrobulbar fat (b
and c, long arrows) typical of idiopathic orbital inflammation.
skull base fracture, orbital or intracranial foreign
body, or traumatic optic neuropathy), contrast mate-
rial adds little to the examination. Finally, for some
disorders such as hydrocephalus, sinus disease
(e.g. sinusitis) or thyroid eye disease (e.g. extraocu-
lar muscle enlargement), contrast material is
unnecessary. In patients with thyroid eye disease,
iodinated contrast may even worsen systemic
thyroid disease.8 When ordering a CT scan it is
important for the ophthalmologist to specify whether
the study should include the orbit, the head or both.
Orbital and head CT scans are widely available, easy
to perform, relatively inexpensive, and the images
can be obtained very rapidly. An orbital CT scan
differs from a head CT in that the images are typi-
cally obtained at a different angle using thinner
slices (e.g. as thin as 0.7 mm vs. 3–4 mm) and can be
reconstructed to any slice thickness (we routinely
reconstruct at 2 mm). For localization purposes, it is
important to view orbital lesions and anatomy in at
least two orthogonal planes. Although the most com-
monly obtained orbital CT planes are axial and
coronal, sagittal views can also be obtained with
computerized reconstruction. Because of the ability
of current generation scanners to obtain very thin
axial slices, the quality of reconstructed coronal (and
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Lee et al.
Figure 1. Thyroid eye disease.
Non-contrast (a) axial and (b)
coronal computed tomography
images demonstrate enlargement
of the medial rectus muscle bellies
compared with those of the lateral
rectus on the axial image, whereas
the coronal image optimally
reveals the true extent of disease
by demonstrating enlargement of
the inferior, medial and superior
rectus muscle bellies bilaterally.
sagittal) images has greatly improved and has obvi-
ated the need for direct coronal scanning at many
institutions. Coronal reconstructions may be the only
option in a patient who is unable to extend or flex
their neck for direct imaging (e.g. trauma, cervical
collar or cervical disc disease). Coronal views
display the anatomical relationships between the
extraocular muscles, optic nerve and surrounding
osseous structures very well. Using the axially
acquired data set, three-dimensional (3-D) recon-
structions can be created (using an independent
workstation), which can dramatically demonstrate
the anatomic relationships between osseous defects
and sinus, orbital or intracranial disease. In the past,
clinicians would implicitly perform a mental recon-
struction of two-dimensional (2-D) images (e.g. axial
and coronal CT images), but 3-D reconstruction
imaging techniques can provide additional informa-
tion for surgical planning or detailed anatomical
analysis.9–11 However, we caution against using 3-D
reconstructions exclusively as the software involved
in producing these images tends to smooth over
abnormalities and may hide subtle pathology, such
as a minimally displaced fracture. The 2-D source
images should always be examined along with the
3-D reconstructions.
© 2008 The Authors
Neuro-ophthalmic and orbital imaging 33

Figure 3. Schwannoma of V1. (a)

Axial and (b) coronal post-contrast
computed tomography images
demonstrate a lesion in the supe-
rior aspect of the left orbit (white
dots). Chronic, non-lytic pressure
changes in the bone of the orbital
roof adjacent to the tumour (b,
arrow) and inferior displacement of
the superior extraocular muscles
(b, double arrows) are identified.

Figure 4. Optic nerve head drusen. Non-contrast axial com-

puted tomography image demonstrates a small calcification at
the left optic nerve head (arrow).
In general, CT is not as sensitive as MR for detect-
ing brain pathology. Although CT scanning is prob-
ably sufficient for many orbital conditions (e.g.
thyroid eye disease [Fig. 1], idiopathic orbital
inflammation [Fig. 2] and orbital tumours [Fig. 3]),
MR is the examination of choice for almost all other
neuro-ophthalmic indications, because it provides
superior soft-tissue discrimination of intracranial
anatomy (e.g. meninges, cavernous sinus, posterior
fossa and dural venous sinuses). However, there are
circumstances where CT may be very helpful, in
addition to MR, such as in demonstrating calcifica-
tion and osseous pathology. Thus, CT is often used in
conjunction with MRI for cases involving disorders
of bone or sinuses (e.g. orbital, maxillofacial, calva-
rial or skull base fracture, hyperostosis of bone asso-
ciated with meningiomas, fibrous dysplasia,
sphenoid wing agenesis in neurofibromatosis-1,
craniosynostosis syndromes, sinusitis or sinus
tumours, clivus or other skull base pathology,
primary bone tumours, bone destruction or erosion).
CT can also be used for demonstrating calcification in
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Figure 5. Craniopharyngioma. Non-contrast axial computed
tomography image demonstrates hyperdense signal intensities
consistent with calcifications (arrow) within this suprasellar mass.
specific ophthalmic conditions (e.g. optic nerve head
drusen12 [Fig. 4], craniopharyngioma [Fig. 5], men-
ingioma or retinoblastoma [Fig. 6]). As described
earlier, a non-contrast CT scan is excellent for
showing the hyperdensity of acute haemorrhage (e.g.
orbital, subdural, subarachnoid, intraventricular or
intraparenchymal haemorrhage associated with
tumour, stroke or ruptured aneurysm). As CT scan-
ning is generally easier to perform, more widely
available, and can be obtained more rapidly than
MR, it is usually preferred ‘after hours’ or in the
emergent setting (e.g. acute trauma, acute stroke,
brain abscess, pituitary apoplexy or intracranial
shunt malfunction), despite MR’s superior soft-
tissue resolution. Some clinical scenarios in which
an ophthalmologist might order an emergent orbital
or head CT scan include the following: (i) acute
orbital trauma (e.g. suspected orbital fracture, orbital
haematoma, metallic or wooden foreign bodies,13,14
34 Lee et al.

Figure 6. Retinoblastoma. Axial

(a) and (b) non-contrast computed
tomography images reveal a large
tumour occupying much of the
right globe. Calcifications, typical
of this tumour, are identified in (b)
(arrow).

Figure 7. Sphenoid wing meningiomas. (a) Axial bone-window computed tomography image demonstrates extensive hyperostosis of
the right anterior clinoid process (arrow) and left greater wing of the sphenoid bone (arrowhead). (b) Axial post-contrast fat-suppressed
T1-weighted magnetic resonance image demonstrates two large meningiomas (arrows). The lesion on the right extends anteriorly to
involve the medial basi-frontal region (white dot). The left tumour involves all sides of the pterion, extending to involve the infratemporal
fossa (black dot) and the lateral, extraconal orbit.

or traumatic optic neuropathy); (ii) acute proptosis
(e.g. orbital cellulitis, idiopathic orbital inflamma-
tion, thyroid eye disease with compressive optic
neuropathy, post-surgical or spontaneous retrobul-
bar haemorrhage); (iii) acute bitemporal hemianop-
sia (e.g. pituitary apoplexy); (iv) acute homonymous
hemianopsia or cortical blindness (e.g. acute stroke,
acute haemorrhage from tumour or arteriovenous
lesion); (v) acute, severe headache (i.e. ‘worst head-
ache of my life’) with or without acute third nerve
palsy (e.g. subarachnoid haemorrhage due to rup-
tured intracranial aneurysm); (vi) acute papil-
loedema (e.g. to rule out intracranial tumour or bleed
in the emergent setting); and (vii) acute visual loss,
headache or diplopia. Finally, there are also relative
and absolute contraindications to MR scanning that
might necessitate the substitution with a CT scan
(e.g. severe claustrophobia, marked obesity exceed-
ing table weight limit, cochlear implant, ferromag-
netic aneurysm clip, pacemaker or other metallic
foreign body). Although the widespread availability
and rapid scan times of CT make it more useful than
MR for patients requiring emergent imaging, a
follow-up MR scan may still be required for optimal
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
evaluation of pathology. Whereas the discussion has
emphasized situations in which one study may be
preferred over the other, CT and MRI are compli-
mentary and can often provide a more complete
picture of the nature of a lesion when used together
than could be found using each separately. This is
particularly true for lesions that have both soft-tissue
and osseous changes (e.g. sphenoid wing menin-
gioma with hyperostosis [Fig. 7] or orbital dermoid
cyst [Fig. 8]).
One major concern with CT scanning is future
cancer risk from the associated low-dose radiation
exposure, particularly in children needing sequen-
tial or multiple imaging studies. Because of the
widespread use of CT, this issue has now become a
public health concern.15 Further work is needed in
this area, but a proposed action plan to address
radiation exposure risk has recently been published
by the American College of Radiology.16,17
Magnetic resonance imaging
Unlike CT, which is based on standard X-ray technol-
ogy, MRI is based on detecting the signal from
© 2008 The Authors
Neuro-ophthalmic and orbital imaging 35

Figure 8. Orbital dermoid. (a) Axial and (b) coronal post-contrast computed tomography images demonstrate a lesion of fat density in
the superior aspect of the right orbit (white dots). Thinning of the adjacent orbital roof is present (b, arrows), indicative of chronicity.
(c) Coronal precontrast T1-weighted magnetic resonance image demonstrates hyperintense signal within the lesion consistent with
fat (arrow). (d) Axial post-contrast fat-suppressed T1-weighted magnetic resonance image confirms the presence of fat by demonstrating
suppression of the fat signal intensity (arrow).

Figure 9. Choroidal melanotic

melanoma. (a) Axial precontrast
T1-weighted magnetic resonance
image demonstrates a round,
homogeneously hyperintense
mass in the posterior–inferior
aspect of the left globe. (b) Coronal
T2-weighted image demonstrates
marked lesion hypointensity. These
signal intensities reflect the para-
magnetic effects of melanin. (Case
courtesy Dr HK Lee.)

© 2008 The Authors

Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
36 Lee et al.

Figure 10. Value of fat

suppression. (a) Axial post-
contrast T1-weighted magnetic
resonance image without fat sup-
pression demonstrates no definite
optic nerve enhancement in this
patient with an inflammatory optic
neuropathy. (b) Axial post-contrast
fat-suppressed T1-weighted image
clearly reveals right optic nerve
enhancement (arrow).

Figure 11. The added value of

fluid attenuation inversion recov-
ery (FLAIR). (a) Axial T2-weighted
magnetic resonance image makes
it difficult to determine the extent
of high signal periventricular white
matter disease because of the
adjacent hyperintense signal of
cerebrospinal fluid (CSF). (b) Axial
T2-weighted FLAIR image demon-
strates suppression of CSF signal
permitting optimal visualization of
the periventricular white matter
lesions (arrows).

Figure 12. Acute intraparenchy-

mal haemorrhage. (a) Axial non-
contrast computed tomography
demonstrates an area of significant
hyperdensity in the left parieto-
occipital region in a patient with
haemophilia A. (b) Axial gradient
recall echo magnetic resonance
image reveals marked signal
hypointensity of this acute
haemorrhage.

© 2008 The Authors

Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
 Table 1. Neuro-ophthalmic indications and recommended imaging study (modified from Lee et al.36 with permission)
Clinical indication
Optic nerve drusen
Bilateral optic disc swelling
© 2008 The Authors
Transient monocular visual
loss (amaurosis fugax) due
to ischaemia
Demyelinating optic neuritis
Inflammatory, infiltrative or
compressive optic neuropathy
Junctional scotoma (i.e. optic
neuropathy in one eye and
superotemporal field loss in
fellow eye)
Bitemporal hemianopsia
Homonymous hemianopsia
Cortical visual loss or visual
association cortex (e.g.
cerebral achromatopsia,
alexia, prosopagnosia,
simultagnosia, optic ataxia,
Balint’s syndrome)
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Third, fourth, sixth nerve palsy
or cavernous sinus syndrome
Internuclear ophthalmoplegia,
supranuclear or nuclear gaze
palsies, dorsal midbrain
syndrome (Fig. 25), skew
deviation
Nystagmus
Hemifacial spasm
Preferred imaging study
CT scan of the orbit may show
calcification (Fig. 4)
MRI head (with MRV)CT scan
might be first-line study in
emergent setting
MRA or CTA of neck for
carotid stenosis or dissection
MRI head and orbit
MRI head and orbit
MRI head (attention to sella)
MRI head (attention to chiasm
and sella) (Figs 23,24)
MRI head
MRI head
MRI head with attention to
the skull base. Isolated
vasculopathic cranial
neuropathies may not require
initial imaging. See Table 3
for third nerve palsy evalua-
tion with MRA or CTA
MRI head (brainstem)
MRI brainstem
MRI brainstem (with or without
MRA)
Contrast material Comment
Not necessary Orbital ultrasound is less costly and more sensitive than a CT scan for
optic disc head drusen12
Yes Consider concomitant contrast MRV to exclude venous sinus
thrombosis (Fig. 19), especially in atypical patients of pseudotu-
mour cerebri who are thin, male or elderly
Depends on clinical situa- Carotid Doppler study might be first line and may still require
tion follow-up catheter angiography
Yes (enhancing lesions FLAIR to look for demyelinating white matter lesions. MRI has
suggest acute disease) prognostic significance for development of multiple sclerosis (Fig. 21)
Yes Fat suppression to exclude intraorbital optic nerve
enhancement (Fig. 22). CT is superior in traumatic optic neuropathy
for canal fractures
Neuro-ophthalmic and orbital imaging
Yes Junctional lesions are typically mass lesions
Yes Consider CT of sella if an emergent scan is needed (e.g. pituitary or
chiasmal apoplexy) or if imaging for calcification (e.g. meningioma
or craniopharyngioma [Fig. 5] or aneurysm)
Yes Retrochiasmal pathway. DWI may be useful if acute ischaemic
infarct (Fig. 13) or PRES (Fig. 14). If structural imaging negative
and organic loss consider functional imaging such as PET
Yes Retrochiasmal pathway. Consider DWI in ischaemic infarct. If structural
imaging negative and organic loss consider functional imaging (e.g.
PET, SPECT or MRS)
Yes Rim calcification in aneurysm, calcification in tumours and hyperostosis
may be better seen on CT
Yes Rule out demyelinating or other brainstem lesion. Include a FLAIR
sequence
Yes Localize nystagmus
Yes Facial nerve compression at root exit zone
37
 38

Table 1. continued

Clinical indication Preferred imaging study Contrast material Comment

Horner syndrome: preganglionic MRI head and neck to second Yes Rule out lateral medullary infarct, brachial plexus injury, apical lung
thoracic vertebra (T2) in neoplasm, carotid dissection, etc.
chest with neck MRA†
Horner syndrome: MRI head and neck to level Yes Rule out carotid dissection. Isolated post-ganglionic lesions are often
post-ganglionic of superior cervical benign
ganglion (C4 level) with
MRA neck‡
Thyroid eye disease CT or MRI of orbit Iodinated contrast may Bone anatomy is better seen on a CT scan especially if orbital
interfere with evalua- decompression is being considered (Fig. 1)
tion and treatment of
systemic thyroid
disease
Orbital cellulitis and orbital CT orbit and sinuses Depends on clinical MRI and/or CT with CTA may be useful adjunct to a CT alone,
disease secondary to sinus situation especially if possible concomitant cavernous sinus thrombosis is
disease present
Idiopathic orbital inflammation CT or MRI of orbit (with fat Yes Beware fat suppression artefacts
suppression)
Orbital tumour (e.g. proptosis CT or MRI of orbit Yes Include head imaging if lesion could extend intracranially. MRI with
or enophthalmos, gaze- contrast is superior at determining intracranial extent of primary
evoked visual loss) optic nerve tumours (e.g. optic nerve glioma or sheath
meningioma) (Figs 26,27). CT scan may be superior if looking for
hyperostosis or calcification
Orbital trauma (e.g. fracture, CT scan of orbit with direct Not generally necessary CT is superior to MRI for bone fractures
subperiosteal haematoma, coronal
orbital foreign body, orbital
emphysema)
Traumatic optic neuropathy CT of optic canal (thin Not generally necessary CT is superior for visualizing fracture or bone fragment
sections)
Carotid-cavernous sinus or dural CT or MRI of head and orbit Yes CT or MRI may show enlarged superior ophthalmic vein. May require
fistula (Fig. 17) (e.g. orbital (with contrast-enhanced catheter angiogram for final diagnosis and therapy. Colour flow
bruit, arterialization of MRA) Doppler studies may be useful for detecting reversal of orbital venous
conjunctival and episcleral flow
vessels, glaucoma)
†
When the presumed lesion is in the lung, mediastinum or anterior aspect of the neck, contrast-enhanced axial CT may be sufficient for localization.37,38 ‡Imaging the entire
pathway (MRI head and neck down to T2 with MRA neck) may be necessary if further localization cannot be performed because of difficulties with hydroxyamphetamine availability.
CT, computed tomography; CTA, computed tomography angiography; DWI, diffusion-weighted imaging; FLAIR, fluid attenuation inversion recovery; MRA, magnetic resonance
angiography; MRI, magnetic resonance imaging; MRS, magnetic resonance spectroscopy; MRV, magnetic resonance venography; PET, positron emission tomography; PRES,
posterior reversible encephalopathy syndrome; SPECT, single photon emission computed tomography.
Lee et al.

© 2008 The Authors

Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Neuro-ophthalmic and orbital imaging
resonance within a large magnetic field. The static
magnetic field created by the MRI scanner is
expressed in the unit Tesla (T). Current MRI scanners
used in clinical practice are typically 1.5 and 3.0 T (up
to 8.0 T on research systems). For comparison, the
earth’s magnetic field is approximately 0.00005 T.18
The MR signal is generated from the interaction of
hydrogen protons (mostly in water) within the pow-
erful magnetic field. Unlike CT, there is no radiation
exposure with MRI. The number of high-field scan-
ners (e.g. 3.0 T) is increasing worldwide because of
the potential technical benefits associated with stron-
ger magnets. These benefits include an increased
signal-to-noise ratio, increased chemical shift resolu-
tion, increased sensitivity to various contrast agents
and improved vessel-versus-tissue contrast. Clini-
cally, this may translate into either higher patient
throughput or improved diagnostic accuracy for many
neuroradiological applications, but this is still
debated.19 There are also proponents of expanding
low-field (<0.5 T) MR technology because of its cost
savings and open format.20 The disadvantages to
using low-field MRI include reduced image quality,
longer scan times and difficulties with contrast
enhancement, fat suppression and DWI.21–24 We
caution against the exclusive use of low-field, open
MRI scanners for orbital and neuroimaging when
high-field scanning is available. The detailed physics
of MRI is beyond the scope of this review and the
reader is referred to other, more in-depth neuroradi-
ology publications for further information.25,26 We
summarize the basic parameters of MRI in the follow-
ing text.
Unlike CT scanning, which relies on the density of
tissue (i.e. hyperdense, isodense or hypodense) to
determine the attenuation of signal, the language of
MR is signal intensity (i.e. hyperintense, isointense
or hypointense). In MRI, the two most common
pulse sequences are T1- and T2-weighted images
(T1WI and T2WI). Proton density imaging will not
be discussed as its use is becoming more and more
limited in orbital and neuroimaging. Normal
anatomy is best demonstrated on T1WI. In distinc-
tion, T2WI are typically better for demonstrating
intracranial or other pathology. Fortunately, the oph-
thalmologist does not need to specifically ask for
T1WI and T2WI as they are typically acquired as part
of the standard MR examination of the orbit and
brain. However, it is important for ophthalmologists
to recognize the MRI appearance of common tissues
in the brain and orbit. The two most important tissue
appearances are the hyperintense T1 signal of fat
(‘bright on T1’), particularly in the orbit, and the
hyperintense T2 signal of cerebrospinal fluid (CSF)
(‘bright on T2’).
Other than fat, there are very few substances that
are hyperintense on precontrast T1WI. Most patho-
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
39
logical lesions tend to be hyperintense on T2WI. For
the ophthalmologist, the most important T1 hyper-
intense substances are subacute haemorrhage (e.g.
pituitary apoplexy), proteinaceous fluid (e.g. sinus
mucoceles or craniopharyngioma) and melanin (e.g.
melanoma).27 Melanin is hyperintense on T1WI and
very hypointense on T2WI, because it is a paramag-
netic substance. This property is helpful when evalu-
ating choroidal melanomas (Fig. 9) or intracranial
melanoma metastases. Hyperintense T2 pathologies
include demyelinating lesions (e.g. periventricular
white matter lesions in multiple sclerosis),
ischaemia (e.g. stroke), inflammatory disease, toxic
or metabolic disorders (e.g. Wernicke encephalopa-
thy) and neoplasms. The ophthalmologist cannot
always rely on fat being hyperintense on T1WI or
CSF being hyperintense on T2WI, because specific
sequences can suppress the normal hyperintense
signal of fat on T1WI (i.e. ‘fat suppression’ or ‘fat
saturation’) or the normal hyperintense signal of CSF
on T2WI (i.e. FLAIR). The ophthalmologist must be
familiar with local scanning protocols to avoid
misdiagnosis. The importance of these suppression
sequences is they eliminate or reduce the hyperin-
tense fat and water signal from normal tissues in
order to enhance visualization of underlying
pathology. Fat suppression sequences allow better
visualization of underlying pathological lesions on
T1WI of the orbit, particularly when contrast is used.
Without fat suppression, the overlying hyperintense
fat signal might obscure underlying contrast
enhancement (e.g. optic neuritis with optic nerve
enhancement [Fig. 10]). Fat suppression techniques
can also confirm the content of fat-containing lesions,
such as orbital dermoid cysts (Fig. 8d) and lipomas.
Although ophthalmologists do not have to order
T1WI or T2WI , they should make sure that fat sup-
pression is used for all post-contrast orbital T1WI
(e.g. for evaluation of optic nerve sheath menin-
gioma or optic neuritis). At many centres, post-
contrast imaging with fat suppression is an
established protocol. Ophthalmologists should be
aware that inadequate or incomplete fat suppression
can be caused by metallic artefacts (e.g. braces or
metallic eyelash mascara) and at the air–bone inter-
face of adjacent paranasal sinuses (e.g. high signal
area at the inferior rectus muscle adjacent to very low
signal air in the maxillary sinus). This hyperintense
signal artefacts could be misinterpreted as abnormal
enhancement and the ophthalmologist should be
suspicious of orbital MR reports describing ‘abnor-
mal enhancement’ in this area (e.g. ‘inflammatory
pseudotumour’ or ‘enhancement in the inferior
rectus muscle’), especially if there is no clinical cor-
relation and there is visible metallic artefacts on the
ipsilateral side. Careful review of other sequences
might confirm that the abnormal signal is indeed
40
Table 2. Guidelines for ordering imaging studies in ophthal-
mology (modified from Lee et al.36 with permission)
1. Decide whether a CT or an MRI scan is indicated or not. The
MRI scan is superior to CT for most neuro-ophthalmic
indications, but CT is superior to MRI for calcification, bone,
acute haemorrhage, if an emergent scan is needed, or if the
patient cannot undergo an MRI scan.
2. Decide if contrast is needed. In most cases, contrast mate-
rial should be ordered for both CT and MRI studies. Contrast
may not be necessary in acute haemorrhage, thyroid eye
disease, or in trauma cases. Caution is necessary for both
iodinated
contrast and gadolinium contrast in patients with renal
failure and contrast may be contraindicated in these
settings.
3. Topographically localize the lesion clinically (‘where is the
lesion’), define the differential diagnosis (‘what is the
lesion’), establish the urgency of the imaging request, and
then order the best study tailored to the lesion
location (e.g. head, orbit or neck).
4. Order specific imaging sequences (e.g. fat suppression for
orbital post-contrast study, fluid attenuation inversion recov-
ery for white matter lesions, gradient recall echo for haem-
orrhage, diffusion-weighted imaging for stroke or PRES)
depending on clinical indication.
5. Order special imaging for specific vascular indications
(e.g. MRA or CTA, MRV, CA). See Table 3 for catheter angiog-
raphy, MRA and CTA recommendations for third nerve palsy.
6. Call the radiologist if there is any doubt about localization,
image study of choice, contrast selection, indications or the
final report.
7. If the imaging shows either no abnormality or an abnormal-
ity that does not match the clinical localization then call the
radiologist or better yet, review the films directly with them.
Ask the radiologist if the area of interest has been
adequately imaged, if artefacts might be obscuring the
lesion, or if
additional studies might show the lesion.
8. If the clinical picture suggests a specific lesion or localization
and initial imaging is ‘normal’ consider repeating the
imaging with thinner slices and higher magnification of the
area of interest especially if the clinical signs and symptoms
are progressive.
9. Recognize that the lack of an imaging abnormality does not
exclude pathology.
CA, conventional angiography; CT, computed tomography;
CTA, computed tomography angiography; MRA, magnetic reso-
nance angiography; MRI, magnetic resonance imaging; MRV,
magnetic resonance venography; PRES, posterior reversible
encephalopathy syndrome.
artefacts.28–31 With regard to CSF intensity, the FLAIR
sequence improves visualization of subtle adjacent
or underlying pathological hyperintensity on T2WI
(e.g. periventricular white matter demyelination
[Fig. 11]).32,33 Ophthalmologists may have to specifi-
cally ask for FLAIR sequences in cases of suspected
demyelinating disease as it may not be routinely
performed as part of a standard MR protocol at some
radiology facilities.
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Lee et al.
Although CT is considered superior to MR for
demonstrating acute haemorrhage (e.g. subarach-
noid haemorrhage), special MR sequences, such as
GRE, can also show blood products effectively. Even
trace amounts of haemorrhage may be detected
because of the profound signal loss caused by the
paramagnetic effects of deoxyhaemoglobin and
methaemoglobin. By using GRE, along with T1WI
and T2WI, the approximate age of a haematoma may
be determined.34 GRE MR sequences are useful for
demonstrating haemorrhage associated with arterio-
venous or cavernous malformations, intra- or extra-
axial intracranial haemorrhage (Fig. 12), or traumatic
brain injury. Evolving GRE techniques, in combina-
tion with FLAIR and T2WI, have also shown
promise in detecting subarachnoid haemorrhage.
However, non-contrast CT still remains the preferred
study for suspected subarachnoid haemorrhage. CT
is faster and less prone to motion artefacts in a dis-
tressed, confused or agitated patient with intracra-
nial haemorrhage.
Gadolinium contrast material in MRI
As with CT, MR scanning with intravenous contrast
material improves detection of pathology by demon-
strating areas of blood–brain barrier breakdown.
Unlike the iodinated contrast used for CT, the con-
trast material used for MR is a paramagnetic material
called gadolinium. The gadolinium paramagnetic
metal ion enhances the local magnetic field and
increases signal intensity. Gadolinium contrast
agents are chelated to form a larger, more stable
complex around the more toxic, free gadolinium.
This stable complex is excreted via the kidneys. For-
tunately, unlike iodinated contrast material, serious
side-effects from gadolinium are uncommon.5 The
most common allergic reactions are typically mild
(e.g. skin rash, sweating, itching, hives and facial
swelling), but there are reports of rare reactions that
can be severe or even fatal. Up until very recently,
MR with gadolinium was considered the ‘procedure
of choice’ in patients with renal insufficiency who
required a contrast study. Unfortunately, a relatively
new disorder related to gadolinium contrast has
emerged called nephrogenic systemic fibrosis (NSF).
28,35
NSF is a disorder seen weeks to months follow-
ing gadolinium contrast administration in patients
with kidney failure. Clinically, NSF is characterized
by thickening and hardening of the skin, typically
over the extremities and trunk. Although most
reports relate to a specific gadolinium agent (i.e.
Omniscan), the Food and Drug Administration has
issued a ‘black box’ warning about NSF and gado-
linium agents approved in the USA (e.g. Magnevist,
MultiHance, ProHance, Omniscan and OptiMark).
Transmetallation (i.e. breaking the bond between
© 2008 The Authors
Neuro-ophthalmic and orbital imaging 41

Figure 13. Acute right occipital

lobe infarction. (a) Diffusion-
weighted imaging demonstrates
restricted diffusion (‘bright’ signal
intensity) in the right occipital lobe
(arrow) in a patient presenting
with a complete left homonymous
hemianopsia. (b) Corresponding
apparent diffusion coefficient map
shows an area of decreased signal
intensity (arrow) confirming an
acute ischaemic stroke.

Figure 14. Posterior reversible

encephalopathy syndrome. (a)
Axial T2-weighted fluid attenua-
tion inversion recovery image
demonstrates hyperintense signal
changes in both occipital lobes in a
hypertensive patient (arrow). (b)
Corresponding diffusion-weighted
imaging reveals iso- and hypoin-
tense (arrow) signal changes con-
sistent with vasogenic oedema.

the toxic metal gadolinium and the chelate), which is
designed to block the toxic, free gadolinium ion,
might be the mechanism for NSF. The American
College of Radiology recommends against using
Omniscan for patients with any stage of renal disease
and caution is advised for all other gadolinium
agents in patients with ‘moderate to severe’ renal
disease. Unfortunately, there is no proven therapy for
NSF.29 The main point for ophthalmologists is that
ordering MRI with gadolinium is no longer a ‘no
risk’ procedure, especially in patients with renal
disease. Despite these new concerns, gadolinium
should probably be ordered for virtually all MR
scans performed for neuro-ophthalmic indications,
unless there is a clear contraindication to the admin-
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
istration of contrast. Gadolinium is very safe in non-
renal failure patients and does not produce cross-
reaction in patients with allergies to iodinated
contrast or fluorescein dye.
Ordering the right imaging study at the
right time
The ophthalmologist should communicate the clini-
cal findings, suspected lesion location, differential
diagnosis and the urgency of the imaging request to
the neuroradiologist on the requisition. Wolintz et al.
described common requisition errors when ordering
ophthalmological studies.30 These authors described
42 Lee et al.

Figure 15. Stroke with perfusion mismatch. (a) Axial T2-weighted fluid attenuation inversion recovery image shows an area of patho-
logical T2 hyperintensity in the right basal ganglia (arrow). (b) Diffusion-weighted imaging (DWI) reveals corresponding restricted diffusion
in the same area as well as a smaller area more posteriorly (arrows). (c) Mean transit time image from a perfusion study demonstrates a
very large area of prolonged transit involving the entire right middle cerebral artery cortical distribution (box) that is not yet completely
infarcted (i.e. the ischaemic penumbra). This represents ‘at-risk’ brain and is consistent with a large DWI/perfusion-weighted imaging
mismatch.

Table 3. Recommended neuroimaging for acute isolated third cranial nerve palsy based on degree of internal and external dysfunction
and risk of aneurysm (modified from Lee et al.63 with permission)

Isolated third nerve palsy Complete external dysfunction Incomplete external dysfunction No external dysfunction
Complete internal Highest risk Highest risk Minimal if any risk
dysfunction MRI with MRA or CTA MRI with MRA or CTA No imaging for
Angiography may still be needed† Angiography may still be needed† aneurysm required
Incomplete internal Uncertain but probably lower risk Moderate risk Minimal if any risk
dysfunction MRI with MRA or CTA MRI with MRA or CTA No imaging for
Consider angiography† Consider angiography† aneurysm required
No internal dysfunction Low risk Uncertain risk Not applicable
Initial observation‡ MRI with MRA or CTA
MRI with MRA or CTA if Consider angiography†
no improvement
†
Catheter angiography should still be considered for patients in whom the risk of aneurysm is higher than the risk of angiography.
Catheter angiography is recommended if (i) worsening of extraocular muscle or iris sphincter impairment continues beyond 14 days; (ii)
iris sphincter impairment progresses to anisocoria >1 mm;65 (iii) no recovery of function occurs within 12 weeks; or (iv) signs of aberrant
regeneration develop.62 ‡Vasculopathic patients (e.g. hypertension or diabetes), with complete external dysfunction and no internal
dysfunction may be observed for improvement without neuroimaging, as it is very likely to be due to an ischaemic third nerve palsy.
Patients without vasculopathic risk factors or vasculopathic patients who do not improve or who progress over a few months time should
undergo a neuroimaging study (MRI with MRA or CTA). At many centres CTA is the superior study for evaluation of aneurysm compared
with MRA. In patients with high suspicion for aneurysm a CTA might be the initial study and then if negative an MRI (with or without an
MRA) might be performed to exclude non-aneurysmal causes for a third nerve palsy. This remains a practice option depending on the
individual center and clinical scenario. MRA, magnetic resonance angiography; MRI, magnetic resonance imaging; CTA, computed
tomography angiography.

the following prescriptive errors: (i) failure to apply a mation; (ii) rejection of a clinical diagnosis because an
dedicated study; (ii) inappropriate use of a dedicated expected imaging abnormality was absent; (iii)
study; (iii) omission of intravenous contrast; and (iv) assumption that a striking imaging abnormality
omission of specialized sequences. The following accounted for the clinical abnormality; and (iv) failure
interpretive errors were also reported: (i) failure to to consider the lack of clinical specificity of imaging
detect the lesion because of misleading clinical infor- abnormalities.30 Table 1 lists the common ophthalmic
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Neuro-ophthalmic and orbital imaging
Figure 16. Posterior communi-
cating artery (PCoA) aneurysm. (a)
Maximum intensity projected
image from a circle of Willis mag-
netic resonance angiography dem-
onstrates a small PCoA aneurysm
(arrow) in a patient with a left
pupil-involving third nerve palsy.
(b) Three-dimensional computed
tomography angiography image
reconstructed on a Vitrea worksta-
tion (same patient) optimally dem-
onstrates the aneurysm sac (white
dot) as well as the aneurysm neck
(arrow).
indications for ordering an imaging study and Table 2
summarizes our recommendations for contacting the
neuroradiologist to discuss the clinical indications,
differential diagnosis, suspected lesion location and
urgency of the study.
Current imaging techniques
Ophthalmologists should be aware of a number of
current MR techniques and sequences that have been
developed to image acute stroke and other central
nervous system (CNS) pathology. These include
DWI, PWI, PCT and constructive interference in the
steady state (CISS).
Diffusion-weighted imaging is an MRI technique
that is based on the microscopic random Brownian
motion of water. These changes in water molecular
diffusion can be measured as signal intensity in vivo
with DWI. In certain pathological states, diffusion
becomes restricted (i.e. hyperintense or ‘bright’ on
DWI). The mobility of water within intracranial
tissue is called the apparent diffusion coefficient
(ADC). The combination of DWI and ADC can define
specific pathology of water diffusion. Thus, DWI can
differentiate the various phases of cerebral infarction:
hyperacute, acute, subacute and chronic. In the
hyperacute phase there is restricted diffusion
(i.e. hyperintense on DWI) and corresponding
hypointense signal changes on ADC, often with
normal T2WI and FLAIR signal intensity. During the
acute phase there is also restricted diffusion and
hypointense signal changes on ADC, but there are
typically hyperintense signal abnormalities on T2WI
and FLAIR images. In the subacute phase, DWI may
still show hyperintensity, but ADC values appear
normal. Finally, in the chronic phase, DWI shows
hypointensity and the corresponding signal changes
on ADC are hyperintense. Thus, DWI can detect
hyperacute ischaemic stroke even before abnormali-
ties are detected on conventional T1WI, T2WI and
FLAIR sequences.
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
43
For the ordering ophthalmologist, the most
common applications for using DWI/ADC are acute
ischaemic homonymous hemianopsia (Fig. 13), corti-
cal visual impairment, evidence of acute cerebral
emboli in a patient with embolic retinal arterial occlu-
sion, or top of the basilar syndrome with acute
brainstem ischaemia producing a progressive oph-
thalmoplegia. DWI and ADC can discriminate
between reversible vasogenic oedema and irrevers-
ible cytotoxic oedema.31,39,40 In vasogenic oedema the
diffusion of water molecules is increased, so the ADC
shows hyperintense and the DWI shows iso- or
hypointense signal. In contrast, in cytotoxic oedema
the movement of water from the extracellular to the
intracellular compartment produces restricted diffu-
sion, so the ADC is hypointense and the DWI is
hyperintense.
One scenario in which the ophthalmologist might
have to decide whether oedema is vasogenic or
cytotoxic in nature is in posterior reversible encepha-
lopathy syndrome. Posterior reversible encepha-
lopathy syndrome typically affects the posterior
brain (e.g. occipital lobes) and can produce
homonymous hemianopsia or cortical visual impair-
ment (Fig. 14).41,42 It is usually encountered in the
setting of hypertension (e.g. eclampsia or amphet-
amine use), but immunosuppressive medications
(e.g. cyclosporine or tacrolimus) can produce a
similar radiographic picture.43 Other DWI/ADC
applications include assessment of inflammatory
(e.g. brain abscess or vasculitis), degenerative (e.g.
Alzheimer dementia), demyelinating (e.g. multiple
sclerosis) and neoplastic lesions (e.g. differentiating
epidermoid from arachnoid cysts).39,44–48
In addition to DWI/ADC, PWI provides additional
information that is useful in assessing the utility of
interventional (catheter) thromblytic therapy for
treatment of acute ischaemic stroke.39,44–50 PWI allows
measurement of capillary perfusion in a region of
interest by signal tracking after bolus injection of a
paramagnetic contrast agent and requires ultrafast
44 Lee et al.

Figure 17. Carotid-cavernous fistula. (a) Axial T1-weighted magnetic resonance image demonstrates a markedly enlarged right superior
ophthalmic vein (SOV) (arrow). (b) Axial and (c) coronal magnetic resonance angiography images also reveal the dilated SOV (b, large arrow)
as well as marked enlargement of the right cavernous sinus (b and c, small arrows). (d) Catheter angiography (same patient) shows the
direct high-flow cartoid cavernous fistula (white dot) draining into the superior ophthalmic vein (large arrow) and inferior petrosal sinus
(small arrow).

Figure 18. Internal carotid artery (ICA) dissection. (a) Axial fat-suppressed T1-weighted magnetic resonance image demonstrates a
normal-appearing flow void in the right ICA (small arrow) and a posterior, hyperintense, crescent-shaped intramural haematoma with a
diminutive anterior flow void in the left ICA (large arrow). (b) Parasagittal T1-weighted image clearly reveals the cranio-caudal extent of the
intramural haematoma (arrows). (c) Anterior and (d) lateral maximum intensity projection images show narrowing of the mid-cervical left
ICA (arrows).

© 2008 The Authors

Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Neuro-ophthalmic and orbital imaging
Figure 19. Venous sinus thrombosis. Contrast-enhanced mag-
netic resonance venography image demonstrates occlusion of
the right transverse (arrow) and sigmoid sinus with extension into
the right internal jugular vein in a patient presenting with bilateral
disc oedema.
MR sequences.51 It provides information about cere-
brovascular haemodynamic parameters, such as cere-
bral blood volume, time to peak, mean transit time
and cerebral blood flow.34 The volume difference of
DWI and PWI (also termed DWI/PWI mismatch)
gives an approximate measure of hypoperfused (but
not yet infarcted), and potentially salvageable tissue
(i.e. the ischaemic penumbra) (Fig. 15). A DWI/PWI
mismatch represents an indication for thrombolysis
based on potential reversibility of DWI lesions, espe-
cially if a vessel occlusion is demonstrated by
MRA.52 For example, if the DWI/PWI mismatch is
greater than 20% and the onset of symptoms is
within a certain time window (e.g. within 6 h for
anterior circulation and within 9 h for posterior cir-
culation), a specific treatment decision for interven-
tion could be made. Perfusion studies are also
becoming an important diagnostic tool in a variety of
other conditions, such as arteriovenous malforma-
tions, CNS neoplasms, epilepsy, and other develop-
mental and degenerative CNS disorders.53
Dynamic perfusion CT imaging is also showing
progress in providing valuable information in acute
stroke and tumour differentiation. Sequential images,
following intravenous injection of iodinated contrast,
are acquired using a CT scanner capable of operating
in the necessary cine mode.54 Good correlation
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
45
between PCT and MR diffusion/perfusion imaging
abnormalities in acute stroke has been demon-
strated.55 Nevertheless, DWI remains the sole imaging
technique with the ability to assess a stroke within
minutes of onset and is unlikely to be replaced by CT
in the near future.56 PCT also shows promise in dis-
tinguishing benign from malignant lesions by deter-
mining tissue perfusion characteristics.57
Finally, recent developments in MR pulse
sequences, such as constructive interference in
steady state (CISS), have allowed higher-resolution
imaging of labyrinthine structures, cranial nerves,
perineural spread of tumour, cavernous sinus inva-
sion and vascular abnormalities. These sequences
have inherent flow compensation that greatly
reduces the artefacts caused by CSF pulsations.58 The
detailed anatomical information provided by CISS
sequences is of particular value in planning surgical
interventions for both intra- and extra-axial lesions.
MRA, CTA, MRV and CTV
Ophthalmologists should also be aware of other
imaging techniques for assessing vascular lesions.
The current widespread use of MRA and CTA has
significantly reduced the use of invasive catheter
angiography for initial evaluation of vascular
pathology.59–61 Digital subtraction catheter angiogra-
phy depicts iodinated contrast-filled vessels without
interference from background tissues. Background
images of the brain are taken as ‘mask images’ and
then ‘subtracted’ from the images after injection of
contrast material. Post-processing, such as window-
ing and filtering, can improve the appearance of the
final images. Technical advances have allowed 3-D
rotational reconstructions that eliminate the need for
multiple oblique views and thereby decrease the
amount of contrast needed, the duration of radiation
exposure and the overall length of the procedure.
These reconstructions also have the advantage of
being better able to delineate the neck of an aneu-
rysm and thereby provide a more accurate assess-
ment of the coil size needed for embolization.15 The
morbidity and mortality of modern catheter angiog-
raphy has been declining with better techniques, but
there remains a significant, albeit small (~1%), risk
of severe complications (e.g. ischaemic stroke).62,63
The technique of MRA relies on the flow-sensitive
nature of the MR signal. On conventional MR, fast-
moving blood appears dark on all sequences as a
‘flow void’. There are two basic types of MRA: (i)
time-of-flight (TOF) MRA; and (ii) phase-contrast
(PC) MRA. Both TOF and PC MRA can acquire data
using either a 2-D or a 3-D scan. The addition of
contrast enhances the MRA acquisition and opti-
mizes visualization of possible dural venous sinus
disease and is also useful for evaluating carotid
46
Figure 21. Multiple sclerosis (MS). Axial T2-weighted fluid
attenuation inversion recovery magnetic resonance image dem-
onstrates multiple, ovoid white matter hyperintense lesions
oriented perpendicular to the lateral ventricles (arrow), charac-
teristic of MS plaques.
stenosis. We believe that TOF MRA is a more robust
technique than PC MRA for evaluating vascular
abnormalities of interest to ophthalmologists (e.g.
aneurysms or arteriovenous malformations). In
either case, the neuroradiologist should examine the
actual source images regardless of which technique
is used.
Magnetic resonance angiography has some advan-
tages over CTA. These include the lack of ionizing
radiation exposure, a less nephrotoxic contrast mate-
rial (i.e. gadolinium vs. iodinated contrast), increased
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Lee et al.
Figure 20. Utility of magnetic
resonance spectroscopy (MRS). (a)
Axial post-contrast fat-suppressed
T1-weighted magnetic resonance
image demonstrates a small right
parietal ring-enhancing lesion
(arrow). (b) MRS shows marked
choline elevation and low N-
acetylaspartate consistent with a
malignant neoplasm. Pathological
diagnosis was a World Health
Organization Grade IV astrocytoma
(glioblastoma multiforme).
Figure 22. Optic neuritis. Axial post-contrast fat-suppressed
T1-weighted magnetic resonance image demonstrates left optic
nerve enhancement (arrow).
signal-to-noise ratio and easier post-processing
techniques. The advantages of CTA over MRA are
increased spatial resolution, a technically easier and
faster study to acquire and less motion artefacts.64
One of the ‘highest stakes’ situations for which the
ophthalmologist might need to decide between inva-
sive (catheter) versus non-invasive (MR or CT)
angiography is in the evaluation of third cranial
nerve palsy. The degree of completeness of external
ophthalmoplegia (external dysfunction) and the
presence or absence of pupil involvement (internal
dysfunction) can be used to determine the initial
imaging strategy for an isolated third nerve palsy.62,63
An imaging approach for assessing an isolated third
nerve palsy is summarized in Table 3 using a risk
tolerance method. For patients with partial or
© 2008 The Authors
Neuro-ophthalmic and orbital imaging 47

Figure 23. Pituitary macroad-

enoma. (a) Sagittal and (b) coronal
post-contrast fat-suppressed T1-
weighted images reveal a large,
homogeneously enhancing sellar
mass with significant suprasellar
extension (arrows) in a patient with
bitemporal hemianopsia.

Figure 24. Craniopharyngioma.

(a) Sagittal and (b) coronal
post-contrast fat-suppressed T1-
weighted images demonstrate a
complex sellar mass with suprasel-
lar extension that contains both a
heterogeneously enhancing solid
component (arrows) and a large
rim-enhancing cystic component
(white dots).

complete external dysfunction of the third cranial
nerve with pupil involvement, MRA or CTA offer
less invasive methods of evaluation than catheter
angiography for detecting aneurysms (Fig. 16). The
advantage of MRA over CTA in the evaluation of a
patient with a third nerve palsy is that an MRI com-
bined with an MRA is likely easier to obtain as both
examinations can be performed using the same
scanner, whereas an MRI combined with a CTA
examination would require two different scanners.
As an MR scan is superior to a CT scan for assessing
non-aneurysmal causes of third nerve palsy, most
ophthalmologists will likely prefer the combination
of MRI/MRA rather than CT/CTA when assessing a
patient with a third nerve palsy.63 The sensitivity and
specificity of both MRA and CTA for ophthalmic
indications are probably equivalent. However, CTA
may be superior to MRA in some instances and at
some institutions, depending on the amount of expe-
rience with this technique. In some cases, MRI with
MRA preceded or followed by CTA may be
necessary. Vasculopathic patients (e.g. hypertension
or diabetes) with complete external dysfunction and
no internal dysfunction (i.e. normal pupil function)
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
may be observed for improvement without neuroim-
aging, as it is very likely to be due to an ischaemic
third nerve palsy. Patients without vasculopathic
risk factors or vasculopathic patients who do not
improve or who progress over a few months time
should undergo a neuroimaging study. The pub-
lished sensitivity of both MRA and CTA for detect-
ing an aneurysm leading to a third nerve palsy is as
high as 98%.62,63,66–68 The size of an aneurysm needed
to produce a third nerve palsy is likely to be within
the detection limit of most scanners used for MRA
and CTA (usually !5 mm in size).62,63 Newer scan-
ners and improved techniques have increased reso-
lution enough to reliably detect aneurysms as small
as 2–3 mm.68 However, as the ‘gold standard’ for
detecting cerebral aneurysms remains catheter
angiography, we recommend its consideration for
highly suspicious cases, even in the face of a negative
MRA or CTA (e.g. poor MRA or CTA visualization of
the entire posterior communicating artery, incom-
plete or poor-quality study, painful pupil-involved
third nerve palsy, other risk factors for aneurysm, or
subarachnoid haemorrhage). As always, the clinician
must weigh the risks of catheter angiography in
48
Figure 25. Pineal germinoma. Axial non-contrast computed
tomography image reveals a hyperdense pineal mass in a patient
with findings of dorsal midbrain syndrome. The normal pineal
calcification is seen in the centre of the mass and the calcified
habenular commissures are seen displaced anteriorly (arrows).
a patient with other significant comorbidities (e.g.
age >70 years, symptomatic atherosclerotic cere-
brovascular disease, significant cardiovascular or
renal disease) against the likelihood of finding an
aneurysm.
For the ophthalmologist, other indications for
ordering a CTA or an MRA include evaluation for
an arteriovenous malformation, dural or carotid-
cavernous fistula (Fig. 17), or suspected carotid artery
disease (e.g. carotid dissection, stenosis or occlu-
sion). In cases of suspected acute arterial dissection,
MRI combined with MRA plays a dominant role in
the initial diagnosis.69,70 Precontrast T1WI may dem-
onstrate a hyperintense, crescent-shaped intramural
haematoma with an eccentric flow void of the patent
lumen at the dissection site. The haematoma can be
more readily distinguished from adjacent fat sur-
rounding the artery with the use of fat suppression
techniques (Fig. 18).71 MRA usually demonstrates
irregularity and narrowing of the artery at the dissec-
tion site. However, it can fail to demonstrate the
dissection if the haematoma is eccentric and does not
cause lumen narrowing. CTA can also be used to
diagnose dissection, especially in the trauma set-
ting.72 When comparing contrast-enhanced MRA
with digital subtraction angiography for evaluation
of carotid stenosis, MRA had a pooled sensitivity of
95% and specificity of 90%.73 CTA has a similar
sensitivity and specificity to MRA, and both have a
slightly higher sensitivity and specificity than
Doppler ultrasonography.74 Multidetector CTA, with
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Lee et al.
its improved imaging capability, has even replaced
conventional catheter angiography in the evaluation
of carotid stenosis at some institutions.74,75 Finally,
the use of MRA or CTA may also be helpful in
patients with hemispheric transient ischaemic
attacks, homonymous hemianopsias, transient
monocular visual loss (i.e. amaurosis fugax), or in
patients with completed cortical strokes.
The major ophthalmic indication for performing a
venogram is to exclude dural venous sinus thrombo-
sis in patients presenting with papilloedema from
increased intracranial pressure (Fig. 19). Symptoms
of cerebral venous sinus thrombosis may mimic
those of idiopathic intracranial hypertension. A stan-
dard head CT can show direct and indirect signs of
venous thrombosis, but in about one-third of cases it
is normal.76 Non-contrast MRI of the brain is more
sensitive than non-contrast CT for showing paren-
chymal abnormalities and the presence of intralumi-
nal thrombus in the sinus. Other MR findings
include the absence of a flow void and the presence
of altered signal intensity in the thrombosed sinus.77
A number of venographic techniques have been
developed to better define sinus anatomy. These
include PC MRV, non-contrast TOF MRV, contrast-
enhanced MRV and CTV. Contrast-enhanced MRV
reduces artefacts that are commonly seen with non-
contrast PC MRV.78–80 Depiction of the dural sinuses
and small-vessel visualization is also improved with
contrast-enhanced MRV compared with TOF MRV.
81,82
An important pitfall on MRV is that flow gaps in
a hypoplastic non-dominant transverse sinus can be
misinterpreted as areas of venous thrombosis. This is
especially problematic on non-contrast TOF MRV. In
such cases, it is important to evaluate the ipsilateral
jugular vein, as it will be smaller on the side of the
hypoplastic sinus, and to correlate findings with the
brain MRI. CTV is another method for detecting
cerebral venous thrombosis, which is both rapid and
widely available.77 CTV can provide very detailed
imaging of the venous system, which is superior to
TOF MRV, and is at least as accurate for detecting
thrombosis.83 Disadvantages to CTV include the use
of iodinated contrast material, ionizing radiation
exposure, and difficulty in reconstructing maximum
intensity projection (MIP) images because of prob-
lems subtracting bone adjacent to the venous
sinuses.77 Our preferred technique for evaluating
cerebral venous sinus thrombosis is a pre- and post-
contrast MRI with a contrast-enhanced MRV.
MR spectroscopy
Magnetic resonance spectroscopy (MRS) is another
evolving MR technique that is based on detecting
various proton MR spectra.84 The commonly mea-
sured metabolites in MRS include N-acetylaspartate
© 2008 The Authors
Neuro-ophthalmic and orbital imaging 49

Figure 26. Optic nerve glioma.

(a) Axial T1-weighted magnetic
resonance image demonstrates
fusiform enlargement and kinking
of the intraconal left optic nerve
(arrow). (b) Axial post-contrast
fat-suppressed T1-weighted image
reveals homogeneous enhance-
ment, as well as involvement of the
proximal optic nerve (arrow).

Figure 27. Optic nerve sheath

meningioma. (a) Axial and (b)
coronal post-contrast fat-sup-
pressed T1-weighted magnetic
resonance images. A pattern of
enhancement of the left optic
nerve sheath is nicely demon-
strated on the axial image (a,
arrows). The coronal image con-
firms the circumferential nature of
the enhancement and optimally
shows the normal-size nerve
centred within the enhancement
(b, arrow), a feature that distin-
guishes this lesion from an optic
nerve glioma.

(NAA), creatine and phosphocreatine (Cr), choline-
containing phospholipids (Cho) and lactate. Other
metabolites, such as glutamate, glutamine, gamma-
aminobutyric acid, myoinositol and fatty acids, can
also be assessed. Although an ophthalmologist is
unlikely to be ordering MRS, a brief description of
the technology is useful. NAA is a neuronal and
axonal integrity marker and a reduction of the NAA
peak on MRS is an indication of neuronal/axonal
dysfunction (e.g. certain neoplasms have no NAA
[e.g. meningiomas] or markedly decreased NAA [e.g.
glioblastoma multiforme {Fig. 20} or metastases]).
Creatine may be elevated in hypometabolic states
(e.g. ischaemia or certain tumours [e.g. gliomatosis
cerebri]) and decreased in hypermetabolic states.
Choline is a component of cell membranes and
increased Cho might suggest increased membrane
synthesis (e.g. actively proliferating or solid, hyper-
cellular tumours). As the Cr level often remains
stable even in disease states, it can be used as a
control for MRS with levels of other metabolites
expressed as a ratio to Cr (e.g. increased Cho/Cr ratio
in certain brain neoplasms). The normal brain
derives energy from aerobic metabolism of glucose
and, therefore, a significant lactate peak on MRS
might indicate anaerobic metabolism (e.g. meta-
bolic disturbances, ischaemia, trauma or brain
neoplasms). Lipid and myoinositol are markers of
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
gliosis and myelin damage and might be elevated in
specific disease states. The spectrum of MRS appli-
cations in neuro-ophthalmology has been expanding
and includes differentiating ischaemic, neoplastic,
demyelinating, radiation necrosis, inflammatory and
mitochondrial disorders.84–87
fMRI, PET and SPECT
As opposed to structural imaging studies (e.g. CT
and MR), functional imaging studies show informa-
tion about the physiology and metabolic function of
brain. These functional studies might be particularly
useful when structural imaging appears normal
despite clinical findings that suggest underlying
brain dysfunction. fMRI is based on changes in T2
signal due to deoxyhaemoglobin.88,89 The signal
change of blood oxygenation level-dependent
mechanisms can be imaged on fMRI. Localized
increases in neuronal activity from task-related brain
activation produce increased cerebral blood flow and
decreased deoxyhaemoglobin and this can be
imaged with fMRI. Presurgical mapping of brain
function, including localization of visual functions,
has provided new insights into brain physiology.
fMRI has many potential advantages because it is
non-invasive, does not require the injection of
contrast agents, and has relatively high spatial and
50
temporal resolution. It can be repeated many times
during the course of a longitudinal study, such as in
clinical drug trials.90,91 There are few clinical indica-
tions for which the general ophthalmologist would
require fMRI.
Other functional imaging studies that have
increasing application in ophthalmology are PET
and SPECT. PET and SPECT both use radiolabelled
molecules to image local metabolic changes (e.g.
regional blood flow and glucose metabolism).92 PET
has superior sensitivity and tissue resolution and is
the preferred study compared with SPECT. 93,94
However, PET is more expensive and less widely
available than SPECT. PET scanning has found
utility in the search for occult neoplasms and moni-
toring patients for tumour recurrence. The use of
PET/CT (combined PET and CT) is now being used,
which allows PET and CT acquisitions during the
same clinical setting without changing the patient’s
position. The combination of the two modalities
improves specificity, as well as sensitivity, in tumour
imaging.95 Less common potential applications for
PET and SPECT include the evaluation of cerebral
toxicity (e.g. cyclosporine-related cerebral blind-
ness), radiation necrosis, stroke, degenerative
disorders, epilepsy, movement disorders and
migraine.93,94,96–98 Except for ocular adnexal lym-
phoma, PET has a limited role in imaging primary
orbital disease.99,100 PET of the orbit is limited by
high signal from the brain and a resolution limit of
7 mm.101,102 Finally, another ophthalmic application
for PET is in assessment of patients with organic
homonymous hemianopsias or cortical blindness
(e.g. carbon monoxide poisoning or visual variant
Alzheimer’s disease).103–105 In these cases the tradi-
tional structural imaging may show no lesion but the
PET might show occipital hypometabolism (‘cold’)
or hypermetabolism (‘hot’).106–109
CONCLUSIONS
Ophthalmologists need to be aware of the basic
mechanics, indications and contraindications of MR
and CT scanning. In general, MRI is superior to CT
for most intracranial neuro-ophthalmic indications.
However, CT still has a role for assessing acute haem-
orrhage, hydrocephalus, osseous pathology, trauma,
orbital disease, sinus disease, thyroid eye disease,
and for evaluating patients unable to have an MR
scan. Neuro-ophthalmic indications and guidelines
for choosing the most appropriate imaging modality
are summarized in Table 1. In general, contrast mate-
rial should be ordered for neuro-ophthalmic indica-
tions, unless there is a clear contraindication.
Awareness of specific MR sequences is needed as
they may not be included in standard imaging pro-
tocols in some regions. These sequences include fat
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Lee et al.
suppression, FLAIR, GRE and DWI. Ophthalmolo-
gists should also be aware of evolving techniques in
vascular imaging (i.e. MRA and CTA) as they are
reducing the need for catheter angiography at many
institutions. Finally, functional imaging, such as PET,
may be indicated in patients with normal structural
neuroimaging studies or for evaluating underlying
systemic inflammatory or neoplastic disorders pro-
ducing eye findings (e.g. paraneoplastic disease).
Once again, we stress the critical importance for the
ordering ophthalmologist to provide the radiologist
with the pertinent clinical findings, a useful differen-
tial diagnosis and the suspected location of a lesion in
order to obtain the best study and interpretation.
REFERENCES
1. Fuchs T, Kachelriess M, Kalender WA. Technical
advances in multi-slice spiral CT. Eur J Radiol 2000;
36: 69–73.
2. Kalender WA, Seissler W, Klotz E, Vock P. Spiral
volumetric CT with single-breath-hold technique,
continuous transport, and continuous scanner
rotation. Radiology 1990; 176: 181–3.
3. Flohr TG, Schaller S, Stierstorfer K, Bruder H, Ohne-
sorge BM, Schoepf UJ. Multi-detector row CT
systems and image-reconstruction techniques. Radiol-
ogy 2005; 235: 756–73.
4. Zagoria RJ. Iodinated contrast agents in
neuroradiology. Neuroimaging Clin N Am 1994; 4: 1–8.
5. Lee AG, Hayman LA, Ross AW. Neuroimaging con-
trast agents in ophthalmology. Surv Ophthalmol 2000;
45: 237–53.
6. Morcos SK, Thomsen HS. Adverse reactions to iodi-
nated contrast media. Eur Radiol 2001; 11: 1267–
75.
7. Smith EE, Rosand J, Greenberg SM. Hemorrhagic
stroke. Neuroimaging Clin N Am 2005; 15: 259–72, ix.
8. van der Molen AJ, Thomsen HS, Morcos SK. Effect of
iodinated contrast media on thyroid function in
adults. Eur Radiol 2004; 14: 902–7.
9. Lancaster JL, Glickman RD, Schulte AV. Strategies for
3-D visualization of ocular structures. Neurosci Biobe-
hav Rev 1993; 17: 451–8.
10. Udupa JK. Three-dimensional visualization and
analysis methodologies: a current perspective. Radio-
graphics 1999; 19: 783–806.
11. John NW, McCloy RF. Navigating and visualizing
three-dimensional data sets. Br J Radiol 2004; 77 (Spec
No. 2): S108–13.
12. Kurz-Levin MM, Landau K. A comparison of imaging
techniques for diagnosing drusen of the optic nerve
head. Arch Ophthalmol 1999; 117: 1045–9.
13. Peterson JJ, Bancroft LW, Kransdorf MJ. Wooden
foreign bodies: imaging appearance. AJR Am J Roent-
genol 2002; 178: 557–62.
14. Yamashita K, Noguchi T, Mihara F et al. An intraor-
bital wooden foreign body: description of a case and
a variety of CT appearances. Emerg Radiol 2007; 14:
41–3.
© 2008 The Authors
Neuro-ophthalmic and orbital imaging
15. Brenner DJ, Hall EJ. Computed tomography – an
increasing source of radiation exposure. N Engl J Med
2007; 357: 2277–84.
16. Amis ES Jr, Butler PF, Applegate KE et al. American
College of Radiology white paper on radiation dose
in medicine. J Am Coll Radiol 2007; 4: 272–84.
17. American College of Radiology. ACR Responds to NEJM
Article on Radiation Risk Associated with CT Scans.
Accessed 19 May 2008. Available from: http://www.
acr.org/MainMenuCategories/media_room/
FeaturedCategories/PressReleases/
ACRRespondstoNEJMArticleonRadiationRisk
AssociatedWithCTScans.aspx
18. Buxton R. Introduction to Functional Magnetic Resonance
Imaging. Cambridge: Cambridge University Press,
2001.
19. Willinek WA, Kuhl CK. 3.0 T neuroimaging: techni-
cal considerations and clinical applications. Neuroim-
aging Clin N Am 2006; 16: 217–28, ix.
20. Hayashi N, Watanabe Y, Masumoto T et al. Utilization
of low-field MR scanners. Magn Reson Med Sci 2004; 3:
27–38.
21. Knauth M, Wirtz CR, Aras N, Sartor K. Low-field
interventional MRI in neurosurgery: finding the right
dose of contrast medium. Neuroradiology 2001; 43:
254–8.
22. Ertl-Wagner BB, Reith W, Sartor K. Low field-low
cost: can low-field magnetic resonance systems
replace high-field magnetic resonance systems in the
diagnostic assessment of multiple sclerosis patients?
Eur Radiol 2001; 11: 1490–4.
23. Delfaut EM, Beltran J, Johnson G, Rousseau J,
Marchandise X, Cotten A. Fat suppression in MR
imaging: techniques and pitfalls. Radiographics 1999;
19: 373–82.
24. Terada H, Gomi T, Harada H et al. Development of
diffusion-weighted image using a 0.3 T open MRI.
J Neuroradiol 2006; 33: 57–61.
25. McRobbie DW, Moore EA, Graves MJ, Prince MR.
MRI from Picture to Proton, 2nd edn. Cambridge: Cam-
bridge University Press, 2006.
26. Weishaupt D, Kochli VD, Marincek B. How Does MRI
Work? An Introduction to the Physics and Function of Mag-
netic Resonance Imaging, 2nd edn. Berlin: Springer,
2008.
27. Caruso RD, Postel GC, McDonald CS, Sherry RG.
High signal on T1-weighted MR images of the
head: a pictorial essay. Clin Imaging 2001; 25: 312–
19.
28. Cowper SE. Nephrogenic systemic fibrosis: a review
and exploration of the role of gadolinium. Adv Derma-
tol 2007; 23: 131–54.
29. US Food and Drug Administration. Gadolinium-
Containing Contrast Agents for Magnetic Resonance Imaging
(MRI): Omniscan, OptiMARK, Magnevist, ProHance, and
MultiHance. Accessed 19 May 2008. Available from:
http://www.fda.gov/cder/drug/advisory/
gadolinium_agents.htm
30. Wolintz RJ, Trobe JD, Cornblath WT, Gebarski SS,
Mark AS, Kolsky MP. Common errors in the use of
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
51
magnetic resonance imaging for neuro-ophthalmic
diagnosis. Surv Ophthalmol 2000; 45: 107–14.
31. Roberts TP, Rowley HA. Diffusion weighted mag-
netic resonance imaging in stroke. Eur J Radiol 2003;
45: 185–94.
32. Arakia Y, Ashikaga R, Fujii K, Nishimura Y, Ueda J,
Fujita N. MR fluid-attenuated inversion recovery
imaging as routine brain T2-weighted imaging. Eur J
Radiol 1999; 32: 136–43.
33. Barkhof F, Scheltens P. Imaging of white matter
lesions. Cerebrovasc Dis 2002; 13 (Suppl. 2): 21–30.
34. Juttler E, Fiebach JB, Schellinger PD. Diagnostic
imaging for acute ischemic stroke management. Expert
Rev Med Devices 2006; 3: 113–26.
35. Issa N, Poggio ED, Fatica RA, Patel R, Ruggieri PM,
Heyka RJ. Nephrogenic systemic fibrosis and its asso-
ciation with gadolinium exposure during MRI. Cleve
Clin J Med 2008; 75: 95–7, 103–4, 106 passim.
36. Lee AG, Brazis PW, Garrity JA, White M. Imaging for
neuro-ophthalmic and orbital disease. Am J Ophthal-
mol 2004; 138: 852–62.
37. Lee JH, Lee HK, Lee DH, Choi CG, Kim SJ, Suh DC.
Neuroimaging strategies for three types of Horner
syndrome with emphasis on anatomic location. AJR
Am J Roentgenol 2007; 188: W74–81.
38. Reede DL, Garcon E, Smoker WR, Kardon R. Hor-
ner’s syndrome: clinical and radiographic evaluation.
Neuroimaging Clin N Am 2008; 18: 369–85.
39. Romano A, Bozzao A, Bonamini M et al. Diffusion-
weighted MR imaging: clinical applications in
neuroradiology. Radiol Med (Torino) 2003; 106: 521–48.
40. Schaefer PW, Copen WA, Lev MH, Gonzalez RG.
Diffusion-weighted imaging in acute stroke. Neuroim-
aging Clin N Am 2005; 15: 503–30, ix–x.
41. Bartynski WS, Boardman JF. Distinct imaging pat-
terns and lesion distribution in posterior reversible
encephalopathy syndrome. AJNR Am J Neuroradiol
2007; 28: 1320–7.
42. Lamy C, Oppenheim C, Meder JF, Mas JL. Neuroim-
aging in posterior reversible encephalopathy
syndrome. J Neuroimaging 2004; 14: 89–96.
43. Cosottini M, Lazzarotti G, Ceravolo R, Michelassi
MC, Canapicchi R, Murri L. Cyclosporine-related
posterior reversible encephalopathy syndrome
(PRES) in non-transplant patient: a case report and
literature review. Eur J Neurol 2003; 10: 461–2.
44. Barboriak DP. Imaging of brain tumors with
diffusion-weighted and diffusion tensor MR
imaging. Magn Reson Imaging Clin N Am 2003; 11: 379–
401.
45. Gregory DG, Pelak VS, Bennett JL. Diffusion-
weighted magnetic resonance imaging and the evalu-
ation of cortical blindness in preeclampsia. Surv
Ophthalmol 2003; 48: 647–50.
46. Kolb SJ, Costello F, Lee AG et al. Distinguishing
ischemic stroke from the stroke-like lesions of
MELAS using apparent diffusion coefficient
mapping. J Neurol Sci 2003; 216: 11–15.
47. Stadnik TW, Demaerel P, Luypaert RR et al. Imaging
tutorial: differential diagnosis of bright lesions on
52
diffusion-weighted MR images. Radiographics 2003;
23: e7.
48. Valentini V, Gaudino S, Spagnolo P, Armenise S,
Tartaglione T, Marano P. Diffusion and perfusion MR
imaging. Rays 2003; 28: 29–43.
49. Bammer R. Basic principles of diffusion-weighted
imaging. Eur J Radiol 2003; 45: 169–84.
50. Rovaris M, Rocca MA, Filippi M. Magnetic
resonance-based techniques for the study and man-
agement of multiple sclerosis. Br Med Bull 2003; 65:
133–44.
51. Rosen BR, Belliveau JW, Vevea JM, Brady TJ. Perfu-
sion imaging with NMR contrast agents. Magn Reson
Med 1990; 14: 249–65.
52. Jansen O, Schellinger P, Fiebach J, Hacke W, Sartor
K. Early recanalisation in acute ischaemic stroke
saves tissue at risk defined by MRI. Lancet 1999; 353:
2036–7.
53. Wolf RL, Detre JA. Clinical neuroimaging using arte-
rial spin-labeled perfusion magnetic resonance
imaging. Neurotherapeutics 2007; 4: 346–59.
54. Wintermark M, Sesay M, Barbier E et al. Comparative
overview of brain perfusion imaging techniques.
J Neuroradiol 2005; 32: 294–314.
55. Eastwood JD, Lev MH, Wintermark M et al. Correla-
tion of early dynamic CT perfusion imaging with
whole-brain MR diffusion and perfusion imaging in
acute hemispheric stroke. AJNR Am J Neuroradiol 2003;
24: 1869–75.
56. Moustafa RR, Baron JC. Imaging the penumbra in
acute stroke. Curr Atheroscler Rep 2006; 8: 281–9.
57. Rumboldt Z, Al-Okaili R, Deveikis JP. Perfusion CT
for head and neck tumors: pilot study. AJNR Am J
Neuroradiol 2005; 26: 1178–85.
58. Glenn LW. Innovations in neuroimaging of skull base
pathology. Otolaryngol Clin North Am 2005; 38: 613–29.
59. Green D, Parker D. CTA and MRA: visualization
without catheterization. Semin Ultrasound CT MR 2003;
24: 185–91.
60. Prince MR, Meaney JF. Expanding role of MR
angiography in clinical practice. Eur Radiol 2006; 16
(Suppl. 2): B3–8.
61. Maldonado TS. What are current preprocedure
imaging requirements for carotid artery stenting and
carotid endarterectomy: have magnetic resonance
angiography and computed tomographic angiogra-
phy made a difference? Semin Vasc Surg 2007; 20: 205–
15.
62. Jacobson DM, Trobe JD. The emerging role of mag-
netic resonance angiography in the management of
patients with third cranial nerve palsy. Am J Ophthal-
mol 1999; 128: 94–6.
63. Lee AG, Hayman LA, Brazis PW. The evaluation of
isolated third nerve palsy revisited: an update on the
evolving role of magnetic resonance, computed
tomography, and catheter angiography. Surv Ophthal-
mol 2002; 47: 137–57.
64. Vaphiades MS, Horton JA. MRA or CTA, that’s the
question. Surv Ophthalmol 2005; 50: 406–10.
65. Jacobson DM. Pupil involvement in patients with
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
Lee et al.
diabetes-associated oculomotor nerve palsy. Arch
Ophthalmol 1998; 116: 723–7.
66. Goldman JP. New techniques and applications for
magnetic resonance angiography. Mt Sinai J Med 2003;
70: 375–85.
67. Lee DH. Magnetic resonance angiography. Adv Neurol
2003; 92: 43–52.
68. Tipper G, U-King-Im JM, Price SJ et al. Detection and
evaluation of intracranial aneurysms with 16-row
multislice CT angiography. Clin Radiol 2005; 60: 565–
72.
69. Oelerich M, Stogbauer F, Kurlemann G, Schul C,
Schuierer G. Craniocervical artery dissection: MR
imaging and MR angiographic findings. Eur Radiol
1999; 9: 1385–91.
70. Paciaroni M, Caso V, Agnelli G. Magnetic resonance
imaging, magnetic resonance and catheter angiogra-
phy for diagnosis of cervical artery dissection. Front
Neurol Neurosci 2005; 20: 102–18.
71. Ozdoba C, Sturzenegger M, Schroth G. Internal
carotid artery dissection: MR imaging features and
clinical-radiologic correlation. Radiology 1996; 199:
191–8.
72. Taschner CA, Leclerc X, Lucas C, Pruvo JP. Computed
tomography angiography for the evaluation of carotid
artery dissections. Front Neurol Neurosci 2005; 20: 119–
28.
73. Nederkoorn PJ, van der Graaf Y, Hunink MG. Duplex
ultrasound and magnetic resonance angiography
compared with digital subtraction angiography in
carotid artery stenosis: a systematic review. Stroke
2003; 34: 1324–32.
74. Koelemay MJ, Nederkoorn PJ, Reitsma JB, Majoie
CB. Systematic review of computed tomographic
angiography for assessment of carotid artery disease.
Stroke 2004; 35: 2306–12.
75. Silvennoinen HM, Ikonen S, Soinne L, Railo M,
Valanne L. CT angiographic analysis of carotid artery
stenosis: comparison of manual assessment, semiau-
tomatic vessel analysis, and digital subtraction
angiography. AJNR Am J Neuroradiol 2007; 28: 97–103.
76. Renowden S. Cerebral venous sinus thrombosis. Eur
Radiol 2004; 14: 215–26.
77. Leach JL, Fortuna RB, Jones BV, Gaskill-Shipley MF.
Imaging of cerebral venous thrombosis: current tech-
niques, spectrum of findings, and diagnostic pitfalls.
Radiographics 2006; 26 (Suppl 1): S19–41; discussion
S42–3.
78. Biousse V, Ameri A, Bousser MG. Isolated intracra-
nial hypertension as the only sign of cerebral venous
thrombosis. Neurology 1999; 53: 1537–42.
79. Friedman DI, Jacobson DM. Diagnostic criteria for
idiopathic intracranial hypertension. Neurology 2002;
59: 1492–5.
80. King JO, Mitchell PJ, Thomson KR, Tress BM.
Manometry combined with cervical puncture in idio-
pathic intracranial hypertension. Neurology 2002; 58:
26–30.
81. Farb RI, Scott JN, Willinsky RA, Montanera WJ,
Wright GA, terBrugge KG. Intracranial venous
© 2008 The Authors
Neuro-ophthalmic and orbital imaging
system: gadolinium-enhanced three-dimensional MR
venography with auto-triggered elliptic centric-
ordered sequence – initial experience. Radiology 2003;
226: 203–9.
82. Farb RI, Vanek I, Scott JN et al. Idiopathic intracranial
hypertension: the prevalence and morphology of
sinovenous stenosis. Neurology 2003; 60: 1418–24.
83. Ozsvath RR, Casey SO, Lustrin ES, Alberico RA, Has-
sankhani A, Patel M. Cerebral venography: compari-
son of CT and MR projection venography. AJR Am J
Roentgenol 1997; 169: 1699–707.
84. Smith JK, Castillo M, Kwock L. MR spectroscopy of
brain tumors. Magn Reson Imaging Clin N Am 2003; 11:
415–29, v–vi.
85. Saunders DE. MR spectroscopy in stroke. Br Med Bull
2000; 56: 334–45.
86. De Stefano N, Filippi M. MR spectroscopy in mul-
tiple sclerosis. J Neuroimaging 2007; 17 (Suppl 1):
S31–5.
87. Lin DD, Crawford TO, Barker PB. Proton MR spec-
troscopy in the diagnostic evaluation of suspected
mitochondrial disease. AJNR Am J Neuroradiol 2003;
24: 33–41.
88. Gore JC. Principles and practice of functional MRI of
the human brain. J Clin Invest 2003; 112: 4–9.
89. Matthews PM, Jezzard P. Functional magnetic reso-
nance imaging. J Neurol Neurosurg Psychiatry 2004; 75:
6–12.
90. Detre JA. Clinical applicability of functional MRI.
J Magn Reson Imaging 2006; 23: 808–15.
91. Dickerson BC. Advances in functional magnetic reso-
nance imaging: technology and clinical applications.
Neurotherapeutics 2007; 4: 360–70.
92. Surti S, Karp JS, Kinahan PE. PET instrumentation.
Radiol Clin North Am 2004; 42: 1003–16, vii.
93. Tai YF, Piccini P. Applications of positron emission
tomography (PET) in neurology. J Neurol Neurosurg
Psychiatry 2004; 75: 669–76.
94. Camargo EE. Brain SPECT in neurology and
psychiatry. J Nucl Med 2001; 42: 611–23.
95. von Schulthess GK, Steinert HC, Hany TF. Integrated
PET/CT: current applications and future directions.
Radiology 2006; 238: 405–22.
96. Powers WJ, Zazulia AR. The use of positron emission
tomography in cerebrovascular disease. Neuroimaging
Clin N Am 2003; 13: 741–58.
97. Hustinx R, Alavi A. SPECT and PET imaging of brain
tumors. Neuroimaging Clin N Am 1999; 9: 751–66.
© 2008 The Authors
Journal compilation © 2008 Royal Australian and New Zealand College of Ophthalmologists
53
98. Hustinx R, Pourdehnad M, Kaschten B, Alavi A. PET
imaging for differentiating recurrent brain tumor
from radiation necrosis. Radiol Clin North Am 2005; 43:
35–47.
99. Valenzuela AA, Allen C, Grimes D, Wong D, Sullivan
TJ. Positron emission tomography in the detection
and staging of ocular adnexal lymphoproliferative
disease. Ophthalmology 2006; 113: 2331–7.
100. Bernardini FP, Bazzan M. Lymphoproliferative
disease of the orbit. Curr Opin Ophthalmol 2007; 18:
398–401.
101. Lane KA, Bilyk JR. Preliminary study of positron
emission tomography in the detection and manage-
ment of orbital malignancy. Ophthal Plast Reconstr Surg
2006; 22: 361–5.
102. Spraul CW, Lang GE, Lang GK. Value of positron
emission tomography in the diagnosis of malignant
ocular tumors. Ophthalmologica 2001; 215: 163–8.
103. Hopkins RO, Woon FL. Neuroimaging, cognitive, and
neurobehavioral outcomes following carbon monox-
ide poisoning. Behav Cogn Neurosci Rev 2006; 5: 141–55.
104. Nestor PJ, Caine D, Fryer TD, Clarke J, Hodges JR.
The topography of metabolic deficits in posterior cor-
tical atrophy (the visual variant of Alzheimer’s
disease) with FDG-PET. J Neurol Neurosurg Psychiatry
2003; 74: 1521–9.
105. Rapoport SI. Positron emission tomography in Alzhe-
imer’s disease in relation to disease pathogenesis: a
critical review. Cerebrovasc Brain Metab Rev 1991; 3:
297–335.
106. Ettl A, Fischer-Klein C, Chemelli A, Daxer A, Felber S.
Nuclear magnetic resonance spectroscopy. Principles
and applications in neuroophthalmology. Int Ophthal-
mol 1994; 18: 171–81.
107. Mathews D, Unwin DH. Quantitative cerebral blood
flow imaging in a patient with the Heidenhain
variant of Creutzfeldt–Jakob disease. Clin Nucl Med
2001; 26: 770–3.
108. Moster ML, Galetta SL, Schatz NJ. Physiologic func-
tional imaging in ‘functional’ visual loss. Surv Oph-
thalmol 1996; 40: 395–9.
109. Uoshima N, Karasuno T, Yagi T et al. Late onset
cyclosporine-induced cerebral blindness with abnor-
mal SPECT imagings in a patient undergoing unre-
lated bone marrow transplantation. Bone Marrow
Transplant 2000; 26: 105–8.