Neuroimaging in neuroophthalmology€¦ · brain imaging evidence of acute intracranial bleeds and contraindications to MR imaging, and 5) osseous abnor-malities including erosion,

Neuroophthalmologist’s Role in Neuroimaging

Although modern neuroimaging modalities play a vitalrole in the management of many neuroophthalmologicaldisorders, they should not be used as a substitute for anexcellent and thorough clinical evaluation.4 Brain MRimages are an integral part of diagnostic criteria or prog-nostic evidence in neuroophthalmological diseases. For ex-ample, a positive MR imaging finding of a plaque lesion is both diagnostic and prognostic in patients with demye-linating optic neuritis. The Optic Neuritis Treatment Trial(ONTT study) concluded that the 10-year risk of MS fol-lowing an initial episode of acute optic neuritis was signifi-cantly higher if there was a single brain MR imaging–docu-mented lesion; higher numbers of lesions did not appreciablyincrease that risk.2 Also, a negative neuroimaging finding isone of the criteria for making a diagnosis of pseudotumorcerebri or idiopathic intracranial hypertension.23 Further-more, live 3D localization of orbital and intracranial tumorshas improved the results of surgery, and functional imaginginvolving PET and fMR imaging has improved our under-standing of the functional basis of disease.

Computed Tomography of the Orbit and Brain

The data points obtained with a CT scanner’s x-ray de-tector are analyzed using a computer, and the scan is repre-sented as pixels of a numerical value resulting from the

attenuation of the x-ray beams. This attenuation coefficient(Hounsfield unit; ranges from 21000 H for air, to 0 H forwater, and to 11000 H for dense bone) provides a numer-ical matrix that helps the computer to reconstruct animage.28

For neuroophthalmologists and orbital surgeons, the clin-ical indications for CT scanning are as follows: 1) orbitaldisorders (thyroid [Fig. 1], trauma, drusen, infection, andtumor), 2) sinus and lacrimal disorders, 3) calcification, 4)brain imaging evidence of acute intracranial bleeds andcontraindications to MR imaging, and 5) osseous abnor-malities including erosion, remodeling, hyperostosis, frac-tures, and calcifications. The disadvantages of CT scanninginclude exposure to radiation, no direct sagittal vantage,possibility of reactions to contrast agent, and dental or os-seous artifacts. The x-ray dose for a standard scan is 3–5rad and 10 rad for a high-resolution scan, which is compa-rable with a standard radiograph dosage. Contrast enhance-ment with iodinated dyes is used to detect intracranial ex-tension of orbital tumors and to evaluate chiasmal andparachiasmal lesions. True coronal slices at intervals of 3mm or less offer the best orientation for examining theorbital contents—to distinguish patterns of extraocular en-largement in thyroid eye disease and orbital pseudotumor,optic nerve lesions such as glioma and nerve sheath menin-gioma, and enlargement of the superior ophthalmic veins inarteriovenous fistulas. Computed tomography has the ad-vantage over MR imaging of requiring less time, being lessexpensive, and being applicable in patients in whom MRimaging is contraindicated or in whom it cannot be per-formed.

Newer-generation scanners employ thinner sections(which may be as small as 1-mm thick) with less volumeaveraging, faster scan times, decreased motion artifact,reduced radiation exposure, more choices of position, and

Neurosurg. Focus / Volume 23 / November, 2007

Neurosurg Focus 23 (5):E9, 2007

Neuroimaging in neuroophthalmology

SWARAJ BOSE, M.D.

Department of Ophthalmology, University of California, Irvine, California

PPRecent advancements in the speed and accuracy of data acquisition and resolution of neuroimaging and interven-tional techniques have revolutionized the early anatomical and functional diagnosis, prognosis, and treatment of manyneuroophthalmological disorders. The relatively new techniques include magnetic resonance (MR) spectroscopy, com-puted tomography angiography, positron emission tomography, and functional MR imaging. In this paper the authordescribes the principles of the current techniques used by neuroophthalmologists and their value in the diagnosis, local-ization, and treatment of various afferent and efferent visual and ocular disorders. (DOI: 10.3171/FOC-07/11/E9)

KEY WORDS • angiography • computed tomography • magnetic resonance imaging •neuroophthalmology

1

Abbreviations used in this paper: AVM = arteriovenous malfor-mation; CT = computed tomography; FDG = 18F-fluoro-2-deoxyglu-cose; fMR = functional magnetic resonance; PET = positron emis-sion tomography; MS = multiple sclerosis; POAG = primaryopen-angle glaucoma; SPECT = single-photon emission CT.

Unauthenticated | Downloaded 07/24/20 11:58 PM UTC

new algorithms that enable sagittal and multiplanar recon-structions. Helical or spiral scanning has reduced acquisi-tion time further because it enables continuous data acqui-sition as the patient table is moved.

Application of Ultrasonography

In B-scan mode, ultrasonography provides a useful inex-pensive, rapid, and easily tolerated adjunct to CT scanningfor orbital disorders because it can distinguish solid fromcystic lesions, show an enlarged superior ophthalmic veinand extraocular muscles, or demonstrate an optic nerve headdrusen.3 Also, transcranial Doppler ultrasonography can beused to assess the patency and location of the temporalartery in patients being evaluated for temporal arteritis.

Magnetic Resonance Tomography and MRAngiography

When body tissue is placed in a strong magnetic field(strength of the magnet that is commercially used variesfrom 1–3 tesla), mobile hydrogen protons align to the mag-netic field; the protons are then exposed to a brief radiofre-quency pulse at a specific (Larmour) frequency that resultsin more protons being antiparallel and, thus, neutralizingmore protons in the opposite direction. The consequence isa decrease in the longitudinal magnetization. The radiofre-quency pulse can also cause the protons to precess in phaseor be synchronous, resulting in a new magnetic vectorcalled the “transverse magnetization.” Computer analysisof the frequency and phase-encoded information from eachslice is converted into spatial localization, and an image iscreated using algorithms similar to those in CT scanning.

The common pulse sequences and techniques used inMR imaging are: spin echo, gradient echo pulse, FLAIR,diffusion weighting, and fat suppression. In FLAIR thesequences help reveal demyelination or MS plaques in thecentral nervous system (Fig. 2), tumors, and ischemic le-sions that often are not visible on routine MR imaging. Thecerebrospinal fluid signal is strongly attenuated, accentuat-ing periventricular and extraaxial disease near the brainsurface.36 Fat suppression is a technique that deletes the fatin the orbit, allowing visibility of small lesions in this loca-tion. Fat-suppression techniques particularly improve thedetection of disease in the orbit, the pituitary gland, andaround the skull base, which has fat in the bone marrow,but also may introduce artifacts, particularly in the loweraspects of the orbit. To measure the phenomenon of slowwater diffusion in tissues, which generally increases in pa-thological states, diffusion weighted images help in theevaluation of cytotoxic edema, demyelinating plaques, in-flammation, tumors, and early brain infarction, as well as todefine internal tissue architecture.8 Acute brain infarction isthe most widely used clinical application of diffusionweighted imaging.

Magnetic resonance angiography does not simply dis-play vascular anatomy, as in contrast angiography. Instead,it extrapolates physiological data obtained from flow char-acteristics of protons to demonstrate anatomy. Thus, in MRangiography, the diameter of the blood vessels sometimesmay appear smaller than that shown on conventional angi-ograms (Fig. 3). Magnetic resonance angiography indica-tions in neuroophthalmology include the evaluation of the

extracranial circulation (carotid artery stenosis, plaques,and dissections in the evaluation of transient visual loss)and the intracranial circulation (aneurysms, AVMs, occlu-sive disease, and carotid artery fistulas). The limitations ofMR angiography are as follows: 1) it cannot detect an-eurysms , 5 mm in diameter, 2) it can yield false-positiveresults in tightly wound vessel loops, and 3) it has a tenden-cy to exaggerate vessel stenosis. Conversely, MR angiog-raphy is an excellent noninvasive technique for detectingasymptomatic aneurysms . 5 mm in size.32

Contrast enhancement is produced using intravenous Gd-diethylenetriamine pentaacetic acid (dose 0.1 mmol/kg), aparamagnetic material that remains extracellular, does notcross the intact blood–brain barrier, and is excreted renally.Because it shortens the T1 relaxation time, Gd typically isused for T1-weighted imaging, in which it provides a brightsignal. For orbital studies, T1-weighted techniques are com-bined with a fat-suppression technique to enhance lesions sothat they may be differentiated from the otherwise brightorbital fat signal. Hemolytic and sickle cell anemia are rela-tive contraindications, whereas rare allergic reactions in-clude hives, bronchospasm, headache, hypotension, or atransient rise in serum iron or bilirubin levels.10

Computed Tomography Angiography

This is a minimally invasive technology that involves anintravenous bolus injection of iodinated contrast, followedby high-speed spiral CT scanning with computer-generated3D images of medium- and large-sized arteries.

The advantages of CT angiography over standard MRangiography are the rapidity of examination and images ofthe true lumen (rather than flow within a vessel), as well asthe fact that it can be performed in patients with claustro-phobia, pacemakers, and older aneurysm clips. The advan-tages of CT angiography include detection of aneurysms assmall as 1.7 mm, superior imaging of the aneurysm neck,better delineation of surgical anatomy, characterization ofmural thrombi, detection of vasospasm, arterial stenosis,and carotid–cavernous fistulas, and provision of rotating3D images. The drawbacks of CT angiography include dif-ficult detection and delineation of cavernous sinus and pos-terior inferior cerebellar artery aneurysms, feeding vesselsfor dural carotid–cavernous fistulas, and risks involvingradiation exposure and contrast agents6,13 (Fig. 4).

Computed Tomography Angiography or MR angiography?

In the treatment of patients with intracranial aneurysmsand especially in the evaluation of patients with painfulpartial cranial nerve III palsy to rule out posterior commu-nicating artery aneurysms, the commonly asked question iswhich imaging modality is superior—CT angiography orMR angiography? (Table 1) Sometimes, the easy answer isthat it depends on the local institutional expertise and theneuroradiologist reading the films. However, clinical datahave shown that contrast-enhanced 3-tesla MR angiogra-phy was comparable in image quality with time of flightMR angiography and CT angiography.42 While other re-ports have shown that a 7-mm posterior communicatingartery aneurysm was missed by brain MR imaging and MRangiography, but detected by CT angiography.41 Further-

S. Bose

2 Neurosurg. Focus / Volume 23 / November, 2007


more, it has been shown that CT angiography was better indetecting traumatic aneurysms in patients with skull basefractures.44 It seems that newer machines have increased inresolution enough to reliably detect aneurysms as small as3 mm.

Conventional Catheter-Based Angiography andAngioembolization

Interventional neuroradiology, currently a fascinatingfield of neuroradiology, provides a relatively safe and reli-able alternative to neurosurgery.43 It involves the introduc-

tion of coaxial systems of extremely flexible microcathe-ters, balloons, coils, chemical agents, and other devices inthe cerebral vascular system for diagnostic and therapeuticpurposes. Common indications are AVMs and carotid–cav-ernous and dural fistulas in which Guglielmi detachablecoils are placed (Target Therapeutics); the risks and com-plications include local hematomas ([# 15%] at the site ofinjection, commonly the femoral artery), vessel wall dis-section (, 1% ), emboli, and transient ischemic attacks orcerebral infarction (1.6% ).14,30

Conventional cerebral angiography remains the goldstandard for accurate detection and localization of smallintracranial aneurysms, in the presence or absence of sub-arachnoid hemorrhage.

Magnetic Resonance Spectroscopy and FunctionalNeuroimaging

Recent advances in neuroimaging technology with PET,SPECT, MR spectroscopy, and fMR imaging have permit-ted new understanding of the neuroanatomical basis of pa-thophysiological phenomena of vision. Current applicationsof functional imaging include detection of hypermetabolicstates associated with tumor, differentiation of tumor fromareas of radiation-induced necrosis, localization of seizurefoci, detection of ischemic regions, evaluation of biochem-ical changes associated with cognitive and psychiatricabnormalities and their response to pharmaceutical inter-vention, and drug localization in the brain.



3

FIG. 3. Magnetic resonance angiogram of the carotid arteries.

FIG. 1. Axial CT scan showing enlarged extraocular muscleswith sparing of the tendons in Graves thyroidopathy, which distin-guishes this lesion from an orbital pseudotumor where the tendonsare involved.

FIG. 2. Brain FLAIR MR image demonstrating the classicperiventricular demyelinating plaques seen in MS.


Magnetic Resonance Spectroscopy

Magnetic resonance spectroscopy is used for diagnosticbiochemistry in vivo and is based on the same principlepreviously used in analytical chemistry to obtain MR spec-tra. With this modality, studies of cerebral ischemia areconfined solely to proton (1H) and 31P because of theirintrinsically higher sensitivity compared with other nuclearspecies. Neurospectroscopy is synonymous with protonMR spectroscopy (higher sensitivity than 31P MR spec-troscopy), which provides results that are easy to interpret,uses small voxel size (1 cm3), and enables the detection of

compounds such as N-acetylaspartate, creatine, choline,lactate, and inositols11 (Fig. 5). Acquiring and interpretingspectra form the basis of a clinical report. Pathological enti-ties documented by MR spectroscopy include brain tumor,stroke, focal cerebral lesions, MS, and intracranial hemor-rhage.20 An acutely ischemic brain produces lactate byanaerobic metabolism during the first 2–3 days after injury,which can be detected by MR spectroscopy. With theadvent of MR spectroscopy and the use of multiple MRmodalities (MR imaging, MR angiography, perfusionweighted MR imaging), it is now possible to evaluateextensively not only regions of cerebral injury but regionsat risk of infarction, and the modalities are very useful inguiding biopsies of brain tumors.38 Another approach tofunctional imaging that correlates MR imaging–delineatedanatomy with magnetoencephalography, which is a map-ping of the magnetic flux, is induced by the background orevoked electrical activity of the brain.

Positron Emission Tomography and SPECT

Positron emission tomography and SPECT are per-formed with systemically administered isotopes (such asFDG, 13NH3, 18F) that emit protons to image biologicalprocesses that measure regional cerebral blood flow andglucose consumption and thus, indirectly, tissue metabo-lism.40 These modalities trace the transport and phosphory-lation of glucose, and the glucose-linked positron emits twophotons, which strike detectors placed around the head.The greater the glucose metabolism of the tissue, the morephotons are emitted. Because FDG cannot diffuse from thebrain, it remains trapped intracellularly and, thus, is anexcellent agent to use for cerebral metabolism imaging27

(Fig. 6). Tomographic images are obtained in a manner similar to

those for MR imaging or CT scanning. Cerebral bloodflow, oxygen utilization, and glucose utilization may bemeasured. For the most part, PET scanning is used for eval-uation of ischemia/stroke, tumors, migraine, blepharo-spasm, cortical visual loss, and mapping of the visual cor-tex, among others.18 The shortcoming of PET is itsrelatively poor resolution of 5–7 mm, high cost, and limit-ed availability because of the requirement for proximity toa cyclotron to produce the radioisotopes.

Positron emission tomography is currently used by vi-sion neuroscientists and researchers to study the anatomicalcorrelates of visual function, and the neuroophthalmolo-gists and orbital surgeons use PET imaging for functionalcorrelates of human diseases including cancers and me-tastatic disease. Richter et al.29 used PET scanning to identi-

S. Bose


TABLE 1Relative risk of an aneurysm developing in a patient with a cranial nerve III palsy and guidelines for neuroimaging*

Extraocular Muscle Palsy

Pupil Total Partial None

pupil involved high (MRA/CTA) highest (MRA/CTA→angio) little/nonenormal pupil lowest (?MRI) low (MRA/close FU) not applicable

* The modifier indicates the relative risk of aneurysm formation and the parenthetical information underscores the optimum detec-tion modality. Note that the risk is highest of an aneurysm in a patient with a pupil involving partial cranial nerve III palsy. Ab-breviations: angio = angiography; CTA = CT angiography; FU = follow-up; MRA = MR angiography.

FIG. 4. Computed tomography angiogram of the carotid arter-ies. Note the simultaneous demonstration of the surrounding os-seous landmarks.


fy the networks involved in the regulation of visual ac-commodation/vergence by contrasting the cortical functionssubservient to eye–lens accommodation with those evokedby foveal fixation. Neural circuits activated selectively dur-ing the near/far response to blur cues over those during con-stant visual fixation occupy posterior structures that includeoccipital visual regions, cerebellar hemispheres and ver-mis, and temporal cortex. We know that color vision is pro-cessed primarily in the ventral stream. In an elegant study,the authors used PET and fMR imaging in humans toinvestigate whether the ventral and dorsal visual streamscooperate when active judgments about color have to bemade. The visual activation sites were identified by retino-topic mapping and cortical flattening. Cortical regions in-volved in dimming detection and motor output includedarea V3A, hMT/V51, lateral occipital sulcus, posteriordorsal intraparietal sulcus, primary motor cortex, and sup-plementary motor area. These experiments demonstratedthat, even with color as the attribute, successive discrimi-nation, in which a decision process has to link visual sig-nals to motor responses, involves both ventral and dorsalvisual stream areas.7

Positron emission tomography can also be useful in the

study of visual pathways involved in amblyopia. It wasused to study the blood flow response in the primary visu-al cortex (V1) to two visual stimuli: low temporal frequen-cy (6 Hz) to activate the parvocellular system and high tem-poral frequency (25 Hz) to activate the magnocellularsystem to investigate pathophysiological mechanism ofamblyopia. The experiments in humans revealed that therewas a decreased activation of blood flow in the contralater-al visual cortex by low temporal frequency stimuli, whichsupports the hypothesis that the parvocellular pathway inamblyopic eyes is depressed.25 This neuroimaging modali-ty has found a great use in the functional evaluation of ma-lignant tumors of the eye (malignant melanoma of the chor-oid, retinoblastoma, metastatic disease, and lymphomas)and brain tumors including pituitary tumors.12,26,34,39

Visual and oculomotor changes are known to occur fol-lowing a stroke. The oculomotor repair during the periodjust after a stroke has been studied using PET.22 Positronemission tomography is also an effective tool to study theeffects of pharmaceutical agents such as neuroprotectants,as well as new and evolving drugs for dementia, Alzheimerdisease, and various optic neuropathies. In one such study,Bose et al.5 found that metabolic imaging with FDG–PET



5

FIG. 5. Results of MR spectroscopy in a patient with a malignant brain lesion with corresponding alteration of thechemical levels. Provided courtesy of Dr. Anton Hasso.


scans demonstrated functional changes in the primary visu-al cortex and visual association areas in all their patientswith nonarteritic ischemic optic neuropathy. Therapy withpentoxifylline for 3 months appeared to reverse or neutral-ize the changes observed in the brain.

In SPECT, isotopes such as 123I-iodoamphetamine or 99Tcare incorporated into biologically active compounds, andthe CT scanner plots their distribution. The informationprovided by SPECT is similar to that of PET, but SPECTdoes not require the use of isotopes produced in acyclotron. However, resolution is even poorer with SPECT.The future of these technologies is very bright, and with theadvent of micro-PET, higher-resolution receptor and genet-ic imaging will provide greater understanding of the work-ings and abnormalities of the human brain.

Functional MR Imaging

Compared with PET and SPECT, fMR imaging is amore current, less invasive technology for mapping cere-bral cortical activation in response to specific cognitive,sensory, or motor tasks performed by an individual. Thebasis for most fMR imaging today is pixel-by-pixel mea-surement of increases in blood oxygenation level duringthe performance of specific tasks (blood oxygen level–dependent imaging)37 (Fig. 7).

The advantage of this technique is that no injection isrequired. Functional MR imaging is evolving rapidly as auseful experimental and clinical tool for functional corticalmapping, psychophysical tests, brain tumor mapping, andunderstanding the basis of higher visual functioning. It hasbeen recently studied to demonstrate the functional andneuroanatomical correlates of visual processing. Visual,oculomotor, and, recently, cognitive functions of the supe-rior colliculi were studied in humans using fMR imaging.This was used to examine activity changes in the humantectum and the lateral geniculate nuclei; blood oxygenlevel–dependent signals in the superior colliculi were com-pared with activity in the inferior colliculi and lateral genic-

ulate nuclei, and the results support a dependency of supe-rior colliculi activity on functions beyond oculomotor con-trol and visual processing.17 Recent anatomical evidenceobtained in nonhuman primates indicates that cingulate

S. Bose


FIG. 7. Brain fMR imaging data obtained while the patient wasobserving a moving target. Note that this modality has better tem-poral resolution, whereas PET (Fig. 6) has superior spatial resolu-tion.

FIG. 6. Three-dimensional surface renderings of mean differences in regional metabolic brain activity (visual cortex)of PET findings. The areas in blue represent a net decrease in regional metabolic activity compared with that noted incontrols.


motor areas play a substantial role in the cortical control ofupper facial movement. Using fMR imaging in humans,the authors concluded that direct cortical innervation of thefacial subnuclei from the cingulate motor areas might con-trol upper face movement in humans, as previously impliedin nonhuman primates.15 Deciding where to look is manda-tory to explore the visual world. Using event-related func-tional MR imaging, the authors found that self-initiation ofsaccades, before their execution, was specifically associat-ed with frontal lobe activation in the dorsolateral prefrontalcortex and in the right presupplementary eye field andfrontal eye fields, suggesting the roles of these areas in thedecision process of where to look when facing two possi-ble visual targets.24 Functional neuroimaging studies of eyemovement control have also been useful for investigatingthe interaction of cognitive and sensorimotor brain sys-tems. Studies of antisaccades, memory-guided saccades,and predictive saccades have helped clarify how cognitivebrain systems support contextually guided and internallygenerated action. Cognitive and sensorimotor eye move-ment paradigms are being used to develop a better under-standing of life span changes in neurocognitive systemsfrom childhood to late life, as well as in behavioral and sys-tems-level brain abnormalities in neuropsychiatric disor-ders.35

Optic neuritis is the first clinical manifestation in approx-imately 20% of patients with MS and is characterized byvisual loss, retrobulbar pain, dyschromatopsia, decreasedcolor and contrast sensitivity, delayed visual evoked poten-tials, and visual field defects. Spontaneous recovery of

vision typically occurs within weeks or months after onset,depending on the resolution of inflammation, remyelina-tion, restoration of conduction in axons that remain demy-elinated, and neuronal plasticity in the cortical and subcor-tical visual pathways. In a study to assess where recoveryoccurs along the visual pathway, visual activation was stud-ied in the lateral geniculate nucleus, the main thalamicrelay nucleus in the visual pathway, and in 3 areas of thevisual cortex (the lateral occipital complexes, V1, and V2)by using fMR imaging. The investigators found that earlycompensatory changes were established in the acute phaseof optic neuritis in the lateral geniculate nucleus and thatthese may indicate very early plasticity of the visual path-ways.21 Despite mapping tools for the central visual field,delineation of peripheral visual field representations in thehuman cortex has remained a challenge. Functional MRimaging mapping has now enabled efficient and robustretinotopic mapping of a wide visual field, which can, atlow cost, be adapted to any clinical environment with visu-al back-projection system.33

Functional MR imaging is a possible means for quanti-fying cortical neurodegeneration in POAG. The functionalorganization of the primary visual cortex (V1) and damageto the optic disc in humans with POAG was measured us-ing the fMR imaging response to a novel method for pro-jecting scotomas onto the flattened cortical representation.The resultant fMR imaging response was compared withinterocular differences in the retinal nerve fiber layer or themean height contour for analogous regions of the visualfield. The authors demonstrated that fMR imaging respons-



7

TABLE 2General guidelines for the choice of neuroimaging modality in neuroophthalmology*

Location Clinical Condition Neuroimaging

orbit tumors USG (solid vs. cystic), CT (noncontrast)MRI (soft tissue, fat suppression)

thyroid ophthalmopathy, trauma, hemorrhage, noncontrast CT (preferred)foreign body

optic nerve tumor, orbital apex tumor Gd-enhanced, fat-suppressed MRI cavernous sinus, chiasm, tumor high-resolution contrast CT (fine cuts), MRI

parasellar region aneurysm (e.g., CN III palsy) Gd-enhanced MRI, MRA, angiographyaneurysm w/ bleeding noncontrast CT

retrochiasmal area & aneurysm or AVM w/ bleed noncontrast CTposterior fossa

brain intracerebral hemorrhage1) acute (intracellular Fe++/metHb) bright† isodense‡ hypodense§2) subacute (extracellular Fe++/metHb) isodense† hyperdense‡ hyperdense§3) chronic (metHb/hemosiderin) dark† hyperdense‡ hyperdense§

papilledema B-scan USG (optic sheath dilation)Gd-enhanced MRI (r/o tumors)MR venogram (venous thrombosis)

MS Gd-enhanced MRI, T2, FLAIR (periventricu-lar plaques)

carotid & vertebral arteries stenosis, dissection, plaques, evaluation of carotid Doppler USG, MRA,CTA, angiogra-amaurosis phy

globe (eyeball) optic disc drusen B-scan USG, noncontrast CTtumor, trauma, calcification noncontrast CT

* Adapted from Bose S: Principles of Imaging in Neuro-Ophthalmology, in Yanoff M, Duker JS (eds): Ophthalmology, ed 2. St.Louis: Mosby, Inc. Abbreviations: CN = cranial nerve; Fe++ = ferrous iron; Fe++/metHb = iron methemoglobin; Fe+++ = ferric iron;MRV = MR venography; USG = ultrasonography.

† Modifier describes density appearance on CT scanning.‡ Modifier describes T1-weighted signal of MR imaging.§ Modifier describes T2-weighted signal of MR imaging.


es to visual stimulation were related to differences in reti-nal nerve fiber layer thickness or mean height contourbetween the eyes. Thus, cortical activity in human V1 wasaltered in these patients with POAG in a manner consistentwith damage to the optic disc.9

Strategies of Imaging in Neuroophthalmology

General guidelines for the choice of neuroimaging mo-dality are summarized in Table 2. Computed tomographyscanning is useful for the evaluation of patients with orbitaldisease (tumor, trauma, and thyroid) and in those with acuteintracranial bleeding. It is important to remember that headCT scans should not be used to interpret orbital disease, andseparate orbital scans should be obtained. Orbital scansrequire negative angulations, parallel to the orbital floor,whereas head scans require positive angulations. Also, theneuroophthalmologist should specifically order coronalviews of orbits (as preferred to reconstructions) in additionto standard axial views for orbital CT scans. Gadolinium-enhanced, fat-suppression MR imaging generally bestdemonstrates diseases of the optic nerve, which includetumors (such as glioma, meningioma, and hemangioma),radiation-induced damage, demyelinating disease, andinflammatory damage (such as sarcoidosis). Papilledemacan be evaluated using B-scan ultrasonography to look fora dilated optic nerve sheath and confirmed by a decrease inthe diameter of the sheath with abduction (or adduction) ofthe eye by 30˚. Brain MR imaging in cases of papilledemawill show slit ventricles, and a different sequence for theveins (MR venogram) may reveal a venous thrombosis.Optic nerve drusen, often calcified, may be seen with B-

scan ultrasonography, CT, MR imaging, or autofluores-cence imaging. For most sellar and parasellar lesions, MRimaging usually is the study of choice. The Optic NeuritisTreatment Trial and the Controlled High-Risk AvonexMultiple Sclerosis study have found MR imaging changesto be of diagnostic and predictive value.1

Combining an understanding of neuroophthalmologicalanatomy with proper imaging techniques provides a pow-erful method to detect lesions involving the afferent andefferent visual pathways. Precise documentation of the ex-tent of injury within the nervous system is becoming in-creasingly important to assess and monitor the effect ofneurological therapies.19 On the contrary, the diagnostic yieldof neuroimaging in patients with normal examination find-ings and isolated, unilateral eye/facial pain referred to a neu-roophthalmologist is low.16 Modern advances in speed andresolution will continue to amaze us as is seen in the 12-teslaMR imaging picture of the postmortem human optic nervethat resembles a histopathological section (Fig. 8).31

References

1. Beck RW, Arrington J, Murtagh FR, Cleary PA, Kaufman DI:Brain magnetic resonance imaging in acute optic neuritis. Ex-perience of the Optic Neuritis Study Group. Arch Neurol 50:841–846, 1993

2. Beck RW, Trobe JD, Moke PS, Gal RL, Xing D, Bhatti MT, et al:High- and low-risk profiles for the development of multiple scle-rosis within 10 years after optic neuritis: experience of the opticneuritis treatment trial. Arch Ophthalmol 121:944–949, 2003

3. Bose S: Orbital tumors, in Levin LA, Arnold AC (eds): Neuro-Ophthalmology: The Practical Guide. New York: ThiemeMedical, 2005, pp 345–355

4. Bose S: Principles of imaging in neuro-ophthalmology, in YanoffM, Duker JS (eds): Ophthalmology, ed 2. St. Louis: Mosby,2004, pp 1241–1248

5. Bose S, Mok AC, Fallon JH, Potkin SG: Three-month change incerebral glucose metabolism in patients with nonarteritic ischemicoptic neuropathy. Graefes Arch Clin Exp Ophthalmol 244:1052–1055, 2006

6. Chow LC, Rubin GD: CT angiography of the arterial system. Ra-diol Clin North Am 40:729–749, 2002

7. Claeys KG, Dupont P, Cornette L, Sunaert S, Van Hecke P, DeSchutter E, et al: Color discrimination involves ventral and dorsalstream visual areas. Cereb Cortex 14:803–822, 2004

8. Doran M, Hajnal JV, Van Bruggen N, King MD, Young IR, Byd-der GM: Normal and abnormal white matter tracts shown by MRimaging using directional diffusion weighted sequences. J Com-put Assist Tomogr 14:865–873, 1990

9. Duncan RO, Sample PA, Weinreb RN, Bowd C, Zangwill LM:Retinotopic organization of primary visual cortex in glaucoma: amethod for comparing cortical function with damage to the opticdisk. Invest Ophthalmol Vis Sci 48:733–744, 2007

10. Elster AD, King JC, Mathews VP, Hamilton CA: Cranial tissues:appearance at gadolinium-enhanced and nonenhanced MR imag-ing with magnetization transfer contrast. Radiology 190:541–546, 1994

11. Ettl A, Fischer-Klein C, Chemelli A, Daxer A, Felber S: Nuclearmagnetic resonance spectroscopy. Principles and applications inneuroophthalmology. Int Ophthalmol 18:171–181, 1994

12. Freudenberg LS, Schueler AO, Beyer T, Antoch G, Kuhl H,Bornfeld N, et al: Whole-body fluorine-18 fluordeoxyglucosepositron emission tomography/computed tomography (FDG-PET/CT) in staging of advanced uveal melanoma. Surv Oph-thalmol 49:537–540, 2004

13. Goddard AJ, Tan G, Becker J: Computed tomography angiogra-

S. Bose


FIG. 8. An imaging “section” of the optic nerve seen using a 12-tesla ultra–high resolution MR imaging system. The arrows pointto the location of the lamina cribrosa.


phy for the detection and characterization of intra-cranial aneu-rysms: current status. Clin Radiol 60:1221–1236, 2005

14. Guglielmi G, Viñuela F, Sepetka I, Macellari V: Electrothrom-bosis of saccular aneurysms via endovascular approach. Part 1:Electrochemical basis, technique, and experimental results. JNeurosurg 75:1–7, 1991

15. Hanakawa T, Dimyan MA, Hallett M: The representation of blink-ing movement in cingulate motor areas: a functional magnetic res-onance imaging study. Cereb Cortex (In press, 2007)

16. Harooni H, Golnik KC, Geddie B, Eggenberger ER, Lee AG: Di-agnostic yield for neuroimaging in patients with unilateral eye orfacial pain. Can J Ophthalmol 40:759–763, 2005

17. Himmelbach M, Erb M, Karnath HO: Activation of superior col-liculi in humans during visual exploration. BMC Neurosci 8:66,2007

18. Hutchinson M, Nakamura T, Moeller JR, Antonini A, BelakhlefA, Dhawan V, et al: The metabolic topography of essentialblepharospasm: a focal dystonia with general implications.Neurology 55:673–677, 2000

19. Jacobs DA, Galetta SL: Neuro-ophthalmology for neuroradiolo-gists. AJNR Am J Neuroradiol 28:3–8, 2007

20. Koopmans RA, Li DK, Zhu G, Allen PS, Penn A, Paty DW:Magnetic resonance spectroscopy of multiple sclerosis: in-vivodetection of myelin breakdown products. Lancet 341:631–632,1993

21. Korsholm K, Madsen KH, Frederiksen JL, Skimminge A, LundTE: Recovery from optic neuritis: an ROI-based analysis of LGNand visual cortical areas. Brain 130:1244–1253, 2007

22. Liebeskind DS, Ances BM, Newberg AB, Jacobs DA, Alavi A:FDG-PET of poststroke oculomotor repair. Neurology 62:1183,2004

23. Mathews MK, Sergott RC, Savino PJ: Pseudotumor cerebri. CurrOpin Ophthalmol 14:364–370, 2003

24. Milea D, Lobel E, Lehéricy S, Leboucher P, Pochon JB, Pierrot-Deseilligny C, et al: Prefrontal cortex is involved in internal deci-sion of forthcoming saccades. Neuroreport 18:1221–1224, 2007

25. Mizoguchi S, Suzuki Y, Kiyosawa M, Mochizuki M, Ishii K: Dif-ferential activation of cerebral blood flow by stimulating ambly-opic and fellow eye. Graefes Arch Clin Exp Ophthalmol 243:576–582, 2005

26. Moll AC, Hoekstra OS, Imhof SM, Coans EF, Schouten-van Mee-teren Ay, van der Valk P, et al: Fluorine-18 fluorodeoxyglucosepositron emission tomography (PET) to detect vital retinoblas-toma in the eye: preliminary experience. Ophthalmic Genet 25:31–35, 2004

27. Pickard JD, Hutchinson PJ, Coles JP, Steiner LA, Johnston AJ,Fryer TD, et al: Imaging of cerebral blood flow and metabolismin brain injury in the ICU. Acta Neurochir Suppl 95:459–464,2005

28. Prokop M: General principles of MDCT. Eur J Radiol 45 (Suppl1):S4–S10, 2003

29. Richter HO, Costello P, Sponheim SR, Lee JT, Pardo JV: Func-tional neuroanatomy of the human near/far response to blur cues:eye-lens accommodation/vergence to point targets varying indepth. Eur J Neurosci 20:2722–2732, 2004

30. Roy D, Raymond J, Bouthillier A, Bojanowski MW, Moumdjian

R, L’Esperance G: Endovascular treatment of ophthalmic seg-ment aneurysms with Guglielmi detachable coils. AJNR Am JNeuroradiol 18:1207–1215, 1997

31. Sadun AA, Carelli V, Bose S, Ross-Cisneros FN, Barboni P, Ah-rens ET: First application of extremely high-resolution magneticresonance imaging to study microscopic features of normal andLHON human optic nerve. Ophthalmology 109:1085–1091,2002

32. Sands MJ, Levitin A: Basics of magnetic resonance imaging.Semin Vasc Surg 17:66–82, 2004

33. Stenbacka L, Vanni S: fMRI of peripheral visual field representa-tion. Clin Neurophysiol 118:1303–1314, 2007

34. Sullivan TJ, Valenzuela AA: Imaging features of ocular adnexallymphoproliferative disease. Eye 20:1189–1195, 2006

35. Sweeney JA, Luna B, Keedy SK, McDowell JE, Clementz BA:fMRI studies of eye movement control: investigating the interac-tion of cognitive and sensorimotor brain systems. Neuroimage 36(2 Suppl):T54–T60, 2007

36. Takanashi J, Sugita K, Fujii K, Takatsuna Y, Niimi H: Optic neu-ritis with silent cerebral lesions: availability of FLAIR sequences.Pediatr Neurol 12:152–154, 1995

37. Thulborn KR: A BOLD move for fMRI. Nat Med 4:155–156,1998

38. Tzika AA, Astrakas L, Cao H, Mintzopoulos D, Andronesi OC,Mindrinos M, et al: Combination of high-resolution magic anglespinning proton magnetic resonance spectroscopy and microscalegenomics to type brain tumor biopsies. Int J Mol Med 20:199–208, 2007

39. Valenzuela AA, Allen C, Grimes D, Wong D, Sullivan TJ: Posi-tron emission tomography in the detection and staging of ocularadnexal lymphoproliferative disease. Ophthalmology 113:2331–2337, 2006

40. van den Hoff J: Principles of quantitative positron emission to-mography. Amino Acids 29:341–353, 2005

41. Vaphiades MS, Horton JA: MRA or CTA, that’s the question.Surv Ophthalmol 50:406–410, 2005

42. Villablanca JP, Nael K, Habibi R, Nael A, Laub G, Finn JP: 3 Tcontrast-enhanced magnetic resonance angiography for evalua-tion of the intracranial arteries: comparison with time-of-flightmagnetic resonance angiography and multislice computed tomog-raphy angiography. Invest Radiol 41:799–805, 2006

43. Wintermark M, Meuli R, Browaeys P, Reichhart M, Bogous-slavsky J, Schnyder P, et al: Comparison of CT perfusion and an-giography and MRI in selecting stroke patients for acute treat-ment. Neurology 68:694–697, 2007

44. Yang YJ, Chen WJ, Zhang Y, Wu ZB, Zhong M, Tan XX, et al:Diagnostic value of CTA and MRA in intracranial traumaticaneurysms. Chin J Traumatol 10:29–33, 2007

Manuscript submitted August 22, 2007.Accepted September 14, 2007.Address correspondence to: Swaraj Bose, M.D., Department of

Ophthalmology, University of California, Irvine, 118 Med Surge I,Irvine, California 92697. email: [email protected].



9


Documents

Neuroimaging in neuroophthalmology€¦ · brain imaging evidence of acute intracranial bleeds and contraindications to MR imaging, and 5) osseous abnor-malities including erosion,