eye movements - jnnp.bmj.com · magneticfield scleral search coil oculographic technique required...

J7ournal ofNeurology, Neurosurgery, and Psychiatry 1995;59:115-125

NEUROLOGICAL INVESTIGATIONS

Eye movements

S Shaunak, E O'Sullivan, C Kennard

Academic Unit ofNeuroscience,Charing Cross andWestminster MedicalSchool, London W68RF, UKS ShaunakE O'SullivanC KennardCorrespondence to:Professor C Kennard,Academic Unit ofNeuroscience, CharingCross Hospital, FulhamPalace Road, London W68RF, UK.

The clinical diagnosis of eye movement disor-ders requires the use of a range of investiga-tions of varying degrees of complexity, fromthe simple cover-uncover test used at thebedside in the evaluation of diplopia to themagnetic field scleral search coil oculographictechnique required to measure accuratelyabnormalities of torsional eye movements.'Before we consider the full range of investiga-tions from the bedside to the laboratory, how-ever, it is essential to understand the use towhich the results of these investigations areput; these include clinical diagnosis, the studyof pathophysiological mechanisms, or deter-mination of therapeutic response. In the caseof diplopia (an awareness of seeing the sameobject in two different locations in visualspace), for example, a systematic approach isrequired to determine which extraocular mus-cles are affected, the aetiology, and the appro-priate management. At each stage it isessential to have a clear understanding of theanatomy and actions of the extraocular mus-cles so that appropriate investigations can beundertaken and the results correctly inter-preted.

Although disturbances of eye position giv-ing rise to diplopia are probably the mostcommon eye movement disorder encounteredby clinical neurologists, evaluation of reflexiveand voluntary eye movements and identifica-tion of nystagmus or some other sponta-neous, involuntary, eye movement canprovide important clues to the topographicaldiagnosis. This is possible because the neuralpathways controlling the different types ofreflex and voluntary eye movements are fairlywell segregated in the neuraxis, at least untilthey feed into the final common "lower"motor neurons in the brainstem, which areresponsible for activation of the extraocularmuscles. In humans the various types of eyemovement all subserve the same goal; the

projection of the image of the object of inter-est on to the most sensitive part of the retina,the fovea. Rapid conjugate eye movements,saccades, enable changes in the line of sightto bring the image of a new object of intereston to the fovea, and the dysjunctive or ver-gence eye movements ensure that theseimages are simultaneously placed on bothfoveae. There is also a need to stabilise theimage of the object of interest on the foveawhen the object itself moves, achieved by thesmooth pursuit system, or when the subject'shead or body moves as occurs during loco-motion when the vestibular and optokineticocular motor reflexes are activated. It is, there-fore, necessary to observe each of these differ-ent types of eye movement in any assessment.We will only briefly mention the vestibularand optokinetic reflexes, as they are fully dis-cussed in the review on balance in this series.

The examination of static eyemovementsACTIONS OF THE EXTRAOCULAR MUSCLESEach eye is rotated by six muscles: four rectiand two obliques (table 1). It should be notedthat the actions of the extraocular muscles aredependent on the starting position of the eye.Hence the superior rectus, because of theanatomy of its insertion into the sclera, actsas a pure elevator only when the globe isabducted by 23°. With increasing adductionof the eye from this position, the superior rec-tus acts more as an intortor and less as an ele-vator. Similarly, the superior oblique actspurely as a depressor only when the eye isadducted, and more as an intortor withincreasing abduction of the eye. To assess thefunction of all the extraocular muscles, eyemovements should therefore be examined inthe nine cardinal positions of gaze.Assessment should include movements of theeyes together, called versions, and of each eyeindividually, called ductions.

Table 1 Action of the extraocular musclesfrom the primary position

Muscle Primary action Secondary action Tertiary action

Medial rectus Adduction -Lateral rectus AbductionSuperior rectus Elevation Intorsion AdductionInferior rectus Depression Extorsion AdductionSuperior oblique Intorsion Depression AbductionInferior oblique Extortion Elevation Abduction

The superior muscles are intortors of the eye and the inferior muscles are extortors.

HERING'S AND SHERRINGTON'S LAWSSherrington's law states that whenever anagonist receives a neural impulse to contract,an equivalent inhibitory impulse is sent to themotor neurons serving the antagonist muscleof the same eye. Every ocular muscle has acontralateral synergist and these muscles, theyoke muscles, act together to move the two

115 on 7 S

eptember 2018 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.59.2.115 on 1 A

ugust 1995. Dow

nloaded from

http://jnnp.bmj.com/

Shaunak, O'Sullivan, Kennard

Table 2 Causes ofincreased square wave jerkfrequency (greater than 15per minute)

Normal elderly subjectsCerebellar diseaseProgressive supranuclear

palsyMultiple system atrophyHuntington's diseaseMotor neuron diseaseSchizophrenia

Table 3 Causes ofmonocular diplopia

Corneal abnormalityIris abnormalityLens abnormalityForeign body (in aqueous or

vitreous humour)Retinal diseaseOccipital cortex pathologyPsychogenic

eyes in the same direction (for example, theright lateral rectus and left medial rectus bothmediate gaze to the right). When an impulseis sent to a muscle causing it to contract, anequal impulse goes to the contralateral syner-gist to maintain parallelism of the visual axes(Hering's law of motor correspondence). Itshould be noted that the fixating eye deter-mines the innervational input to both eyes.This is of importance in the assessment of thecover test, and also in the interpretation ofinvestigations such as the Hess screen test.

ABNORMALITIES OF FIXATIONThe subject's eyes should be observed in theprimary position while fixing an object thatrequires visual discrimination. Each eyeshould then be occluded in turn, and anyabnormalities such as latent nystagmusobserved. The use of an ophthalmoscope toview one optic disc while the patient fixateswith the other eye will magnify any move-ments seen. The most common intrusions aresquare wave jerks, which have an amplitudeof 0.5-10' with refixation within 250 ms. Ahigh frequency of square wave jerks has beenreported in cerebellar disease2 and parkinson-ism plus syndromes,3 among other conditions(table 2).

ASSESSMENT OF DIPLOPIADiplopia is among the most commonlyencountered neuro-ophthalmological symp-toms in neurological practice, and usuallyarises from a disparity in retinal stimulationbetween the two eyes. If diplopia is presentwith one eye covered, an optical aberrationwithin the refracting media of the eye is likelyto be present, although there are other causesof this phenomenon.8 (table 3). If diplopia isalleviated by covering one eye, a systematicapproach to evaluation is required. As well asdetermining the nature of separation of thetwo images and the direction of maximal sep-aration, enquiries as to the presence of a fam-ily history of strabismus, or a childhoodhistory of orthoptic treatment should bemade. If the eyes are misaligned, it should beascertained at an early stage if one is dealingwith a non-comitant or comitant strabismus;the degree of misalignment varies with gazeposition in the first, but does not vary withgaze position in the second. Non-comitancesuggests a recent paretic or restrictive aetiol-ogy. Comitance is characteristic of childhoodstrabismus, and diplopia in such circum-stances is usually due to decompensation of alongstanding phoria (a deviation of the visualaxes when only one eye is viewing, normallykept in check by fusional mechanisms-thatis, a latent deviation). The term tropia as usedlater refers to a deviation of the visual axeswhen both eyes are viewing, which is not keptin check by fusion (a manifest deviation).

Head posturePatients with diplopia may adopt a compen-satory head posture, and the position of thechin, head, and face should therefore be care-fully observed. The purpose of an abnormal

head posture is to turn the eyes as far as pos-sible from the field of action of the weak mus-cle. Hence, if one of the muscles thatmediates conjugate gaze to the left is under-acting, the face will also be turned to the left.Underaction of the superior or inferior recti,which act primarily to move the eyes in thevertical plane, is compensated by head flexionor extension respectively. Torsional diplopiausually arises from underaction of thesuperior or inferior oblique muscles, andpatients with this symptom often tilt theirhead towards the shoulder opposite to that ofthe weak muscle.

Hirschberg's testThis is a rough objective test to determine thedegree of tropia, and is particularly useful inyoung or uncooperative patients. The cornealreflections should be observed while thepatient fixates a light source at a distance of33 cm with both eyes open, and then whileeach eye fixates in turn. A decentration of thecorneal reflex by 1 mm corresponds to about70 of ocular deviation. The Krimsky test is avariation of this in which prisms are placed infront of the deviated eye until the cornealreflections are symmetric. It should be notedthat a small angle strabismus will displacecorneal reflections by a degree that is unlikelyto be clinically detected.

Ocular ductions and versionsOcular ductions and versions should beassessed in the nine cardinal positions ofgaze. Ductions may not show minimal muscleweakness that can be overcome by thepatient, but versions will often show a subtleparesis. A limitation of movement in a partic-ular direction may be related to paresis of theagonist muscle or tethering of the ipsilateralantagonist muscle. Apparent underaction of amuscle may also arise from the phenomenontermed inhibition of the contralateral antagonist.This arises when a patient fixes with theparetic eye. In these circumstances the unop-posed ipsilateral antagonist of the pareticmuscle requires less neural input than normalto move the eye. Consequently, fromHering's law, the contralateral yoke musclealso receives subnormal innervation, andseems to have limited excursion. This mostcommonly causes confusion between trueparesis of a superior oblique muscle andapparent paresis of the contralateral superiorrectus. In such a situation, however, ocularductions will be full in the non-paretic eye,although ocular versions may suggest limita-tion of movement of the contralateralsuperior rectus if the paretic eye is fixating.

IDENTIFICATION OF THE PARETIC MUSCLECover-uncover test (fig 1A)Cover tests rely on the fact that foveationoccurs in an eye that is forced to fixate. If theretinal image was not directed on to the foveabefore the eye took up fixation, a movementof redress will be noted as the eye fixates,which gives an indication of the degree ofmisalignment of the visual axes.

116 on 7 S



j.com/

J Neurol N


ugust 1995. Dow

nloaded from


Eye movements

A Cover-uncover test

(a) R L (a) R

B Alternate cover test

L (c) R L

-*-=Secondary Primarydeviation deviation

(d)(b)

0,,,,,,,,,

Figure 1 (A) Cover-uncover test, showing the presence of an esophoria. Dotted lines indicate the position of the eye whenunder cover. (a) At rest the visual axes are aligned correctly. (b) When the cover is placed before the left eye, the eye no

longerfixates, and moves inwards. (c) On removal of the cover the eye moves outwards to take up fixation, indicating anesophoria. (B) The alternate cover test, showing the presence of an esotropia. (a) At rest, with both eyes viewing, there is amanifest inward deviation of the left eye. (b) A cover placed before the non-fixating left eye causes no movement. (c)When the right eye is occluded, the left eye is forced to fixate, and a movement of redress occurs (the primary deviation).The resulting additional innervation to the contralateralyoke muscle leads to deviation of the sound eye under the cover

(the secondary deviation). Note that the secondary deviation is greater than the primary deviation. (d) When the cover istransferred to the left eye both eyes assume their original position.

The cover-uncover test should be per-formed both before and after the correctionof any abnormal head posture, and with theeyes in the nine cardinal positions of gaze. Aclearly defined fixation target at a distance of6 m should be used, and the test repeatedwith a near target at a distance of 33 cm todetermine the effect of vergence and accom-

modation on any response seen. The test isperformed by occluding one eye at a time,and initially observing the movements of theuncovered eye. If the uncovered eye moves totake up fixation, it can be assumed that underbinocular conditions the eye was not alignedwith fixation, and a manifest deviation was

present (a tropia). Inward movement of theuncovered eye indicates an exotropia, and out-ward movement an esotropia. A vertical devia-tion may be either a hypotropia or a

hypertropia, depending on whether the eyemoves up or down respectively. The exam-iner should determine whether the tropia iscomitant or non-comitant by seeing if themagnitude of the deviation varies in the dif-ferent positions of gaze. If no tropia is pre-sent, and the covered eye is noted to move toassume fixation just after it is uncovered, a

latent deviation (a heterophoria) is present,and this may also be classified as an exopho-ria, esophoria, hypophoria, or a hyperphoriadepending on the direction of the deviation.The test is then repeated, and the same

observations made while covering the othereye. It should be noted that the convention isthat if there is a vertical deviation of the eyes,the higher of the two is referred to as hyper-tropic/hyperphoric, regardless of which eye isactually at fault.

Alternate cover test (fig IB)The alternate cover test is more dissociatingthan the cover-uncover test, and should beused to fully dissociate the eyes and show themaximal deviation. While the patient fixates a

target the occluder is quickly switched fromeye to eye to prevent binocular viewing,allowing sufficient time for the eyes to settlein their new position. The alternate cover testshould also be performed in the nine cardinalpositions of gaze to determine the direction ofgaze that elicits the maximal deviation, andthe eye in which fixation in that field of gazecauses the greater deviation. It is importantthat the patient should never be allowed toregain fusion while the occluder is beingtransferred. The examiner should note themovement of the uncovered eye as theoccluder is changed from one eye to theother. Movement of the uncovered eye mayindicate either a heterotropia or a heteropho-ria, and the alternate cover test will not differ-entiate between these possibilities. Thecover-uncover test must therefore be per-formed first to determine if a tropia is pre-sent.The size of the deviation of the paretic eye

when under cover, with the non-paretic eyefixating, is termed the primary deviation.When the paretic eye is forced to fixate, addi-tional innervation is required to overcome theparesis. This excessive innervation is also, byHering's law, equally transmitted to the con-

tralateral synergist, which consequently over-acts. This overaction is termed the secondarydeviation, and is greater than the primarydeviation. The alternate cover test best showsthe difference in size between the primary

(b)

0////i////(c)

117

.4

on 7 Septem

ber 2018 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.59.2.115 on 1 August 1995. D

ownloaded from



and secondary deviation, and this finding canbe used to identify the weak muscle in a yokepair by comparing the movements of redressin the two eyes. It should be noted that themovement of redress is equal in both eyes inthe case of comitant strabismus.

Prisms can be used in conjunction with thealternate cover test (prism and cover test) toquantify the deviation by determining theprism strength that nullifies or just reversesthe movement of redress. This test measuresthe total deviation and does not separatetropia from phoria.

The red glass testThis technique allows the patient withdiplopia to readily differentiate between theimages from each eye. A red filter is placed infront of one eye (by convention the right) sothat the subject with diplopia sees two differ-ent coloured images. The separation of theseimages can then be reported by the patient inthe nine cardinal positions of gaze, and inhead tilts to the right and left. The red imageis towards the right (uncrossed) with anesotropia, and towards the left (crossed) withan exotropia. The separation of the twoimages becomes maximal in the direction ofgaze of the paretic muscle, with the imagefrom the paretic eye projected more peripher-ally.

The Maddox rod testThis technique provides a subjective methodof measuring ocular deviations and dependson dissociation of the eyes by presenting apoint source of light to one eye and a lineimage to the other eye. The Maddox roddevice consists of small glass rods with a redfilter, and has the effect of transforming apoint source of light into a line perpendicularto the axes of the cylinder; hence the rodsmay be oriented according to the desiredplane of testing. The test is particularly usefulin torsional diplopia where superior or infe-rior oblique muscle weakness is suspected. Inthe case of a suspected right superior obliquepalsy, for example, the rod is held horizon-tally before the right eye while the left eyeviews a white point source of light. Thepatient will report that the red line is lowerthan the point source, and relatively intorted,and the separation of the images will be maxi-mal on looking down and to the left. By rotat-ing the rods in their frame until the red line isvertical the amount of cyclotropia (torsionaldeviation) can be determined. A variation ofthe test uses both red and white lenses so thatthe subject can compare the position and ori-entation of two lines, rather than a line andpoint source.The Maddox rod test primarily detects the

presence of phorias, because the red lens usedproduces very dissimilar images and thereforeprevents fusional vergence. It should be notedthat some patients may be able to fuse thetwo images in the red glass test, and a phoriacan therefore be overlooked. For this reason,the results of the Maddox rod test are usuallymore reliable. The clinician should be aware

that orthophoria (the condition of perfectalignment of the visual axes) is a physiologi-cally unusual state, and normal people oftenhave a small comitant phoria.

Parks-Bielschowsky testfor vertical diplopiaThe Parks-Bielschowsky test is used to ascer-tain the weak muscle in patients with verticalor torsional diplopia. The first step is todetermine with the cover-uncover testwhether there is a right or left hypertropia inthe primary position and after correction ofany abnormal head posture. A right hyper-tropia, for example, may arise from underac-tion of the depressors of the right eye (rightinferior rectus and superior oblique).Alternatively, the left eye may be hypotropicbecause of weakness of the elevators of thateye (left inferior oblique and superior rectus).The alternate cover test should then be usedto determine whether the amount of verticaldeviation increases in right or left gaze. If thehypertropia increases on left gaze, the rightsuperior oblique or left superior rectus areunderacting. Further differentiation betweenthese alternatives is possible by asking thesubject to look up and down in left gaze. Anincrease in hypertropia in gaze downwardsimplicates the superior oblique as the weakmuscle.

Finally, the vertical deviation should becompared with the alternate cover test inright and left head tilt positions. The degreeof misalignment will increase when the headis tilted to the side of the paretic muscle if theipsilateral intortors (superior oblique andsuperior rectus) are weak, and to the oppositeside if the extorting muscles (inferior obliqueand inferior rectus) are weak. In practice, anincreased misalignment on head tilt is usuallyindicative of an ipsilateral superior obliquepalsy, although the test will help to differenti-ate this from apparent weakness of the con-tralateral superior rectus arising from thephenomenon of inhibition of the contralateralantagonist (see earlier). The test is less oftenpositive with palsies of the vertical recti orinferior oblique muscles.The explanation for the effect lies in the

fact that a head tilt to either shoulder inducesan ocular counter rolling, which is mediatedby the ipsilateral intortors (superior rectusand superior oblique), and the contralateralextortors (inferior rectus and inferioroblique). If, for example, the ipsilateral supe-rior oblique is paretic, the superior rectus onthe same side receives excessive innervationto intort the eye, and by virtue of its relativelyunopposed primary action elevates the eye.

The Hess screen testThis test is used in the investigation of non-comitant strabismus to assess the paretic ele-ment, and depends on the use of mirrors orfilters to dissociate the eyes and show theposition of the non-fixing eye when the othereye is fixing in specified positions of gaze.9The test is invaluable in providing a perma-nent record of ocular motility that can beused to monitor progress and treatment. Two

118 on 7 S



j.com/

J Neurol N


ugust 1995. Dow

nloaded from


Eye movements

test objects are presented in the field of viewand the patient is required to place them insuch a fashion that they seem to be superim-posed (the haploscopic principle).

In the Lee's screen adaptation of the test,the apparatus consists of two glass screens at900 to each other, which are bisected by adouble sided plane mirror. A grid on eachscreen is marked with dots at 50 intervals,which are connected to form inner and outerfields at displacements of 150 and 300 respec-tively. The subject fixates with one eyethrough the mirror the dots on one of thescreens which is illuminated, and the exam-iner uses a pointer to indicate the position ofa dot in the field of this eye. The patient isrequired to place his pointer on the non-illu-minated screen, viewed with the other eye, sothat it is superimposed over the foveal imageof the fixating eye. The blank screen is thenbriefly illuminated, and the position of thepointer recorded on a chart that is a copy ofthe grid on the screen. As the innervation ofthe extraocular muscles of both eyes is deter-mined by the fixating eye, any muscle under-action or overaction in the non-fixating eyecan be identified. The procedure is per-formed in the cardinal positions of gaze, andthen repeated with the other eye fixating.The charts from the two eyes are assessed

by comparing the plotted fields with eachother and with the normal fields on the chart.A difference in the size of the fields showsnon-comitance, which usually indicates therecent onset of paresis. The smaller of thetwo fields indicates the primarily affected eye.If the fields are displaced but of the same size,comitance is present, and this suggests alongstanding deviation or a non-paretic aeti-ology.

Each field is then compared with the nor-mal field. The position of the central dot inthe smaller field indicates the primary devia-tion (the result of fixating with the unaffectedeye), and its position in the larger field indi-cates the secondary deviation (the result offixating with the affected eye). Underaction isidentified as inward displacement of the dots,and overaction as outward displacement;maximum displacement will occur in thedirection of action of the overacting contralat-eral synergist. A narrow, symmetric field withrestriction in opposite directions implies amechanical restriction of ocular movement.The outer field should also always be exam-ined; this may show a defect when the centralfields appear normal, particularly when amechanical defect is present, or in cases ofslight paresis.

Secondary changes may occur with timethat make determination of the primarilyparetic muscle difficult (spread of comitance),and the Hess chart in such circumstances willbecome increasingly comitant. As well asoveraction of the contralateral synergist, over-action of the ipsilateral antagonist due to con-tracture, and inhibition of the contralateralantagonist will appreciably alter the appear-ance of the Hess chart. In these circum-stances the fact that the overaction of the

contralateral synergist to the primarily pareticmuscle remains slightly greater than that ofthe ipsilateral antagonist may help to identifythe muscle that was initially paretic.

The Lancaster red-green testThe Lancaster red-green test makes use ofthe same basic principles, but uses red andgreen filters to dissociate the eyes.'0 Thepatient wears reversible goggles with a red fil-ter in front of the right eye and a green filterin front of the left eye, and therefore sees onlythe image of a red light with the right eye,and the image of a green light with the lefteye. The examiner projects the linear imageof a red torch on to a screen in the nine cardi-nal positions of gaze and the patient is askedto superimpose the image of a green torch onto the screen so that the two images seem tohim to exactly coincide. If there is a deviationof the visual axes the two images will be sepa-rated on the screen. The effect of fixatingwith either eye can be investigated by eitherreversing the goggles, or simply by the patientand examiner exchanging torches.

RESTRICTIVE MUSCLE DISEASEForced duction testingThe forced duction test is used to determinethe presence of a mechanical restriction toeye movements. This may occur in, for exam-ple, thyroid ophthalmopathy," Brown's supe-rior oblique tendon sheath syndrome,'2 orafter orbital blowout fractures (table 4). Localor general anaesthesia is required. Two pairsof forceps are applied to diametricallyopposed limbal points, and horizontal,oblique, and vertical rotations of the globe areperformed. Failure of movement or retractionof the globe into the orbit indicates a restric-tive process. Care must be taken that theglobe is not inadvertently pushed back intothe orbit, as this increases the relative lengthof the tethered muscle, and may mask a posi-tive result.An alternative, less invasive, procedure,

which may also be used in an uncooperativepatient, is to measure intraocular pressures inthe primary position and again with gaze inthe direction of the limitation of movement.An increase in pressure of greater than 6 mmHg suggests a mechanical limitation.

THE TENSILON TESTMyasthenia gravis may mimic any single orcombined extraocular muscle palsy, orinfranuclear or supranuclear ophthalmople-gia, and must therefore be considered in thedifferential diagnosis of any puzzling acquiredocular motility disturbance. Although the

Table 4 Causes of restrictive ophthalmopathy

Thyroid diseaseBrown's syndrome: congenital or acquiredDuane's syndromeBlow-out fracturesInfiltration: for example, metastases, lymphoma, amyloid,orbital myositis, pseudotumourCaroticocavernous fistulaExtraocular muscle fibrosisProlonged muscle weakness with secondary contracture

119 on 7 S



j.com/

J Neurol N


ugust 1995. Dow

nloaded from



Tensilon test remains an important tool inthe diagnosis of this condition, the clinicianmust be aware of the limitations and potentialpitfalls of the procedure.'4 In particular, it isimportant that a definite goal or end point bechosen such as a quantifiable change in thedegree of ptosis or ocular misalignment (forexample, pretest and post-test Hess charts orcover and prism tests).

Bedside examination of dynamic eyemovementsAfter examination of static eye movementsthe various dynamic eye movement systems,including saccades, pursuit, vestibulo-ocularreflex (VOR), and optokinetic nystagmus(OKN) should be tested, firstly at the bedsideand then if necessary by oculography. Itshould be noted that oculomotor perfor-mance may be modulated by the patients'age, attentiveness, and concurrent medica-tion.

SACCADESVoluntary saccade initiation should beassessed by instructing the patient to lookfrom side to side. The patient is then asked tofixate two targets alternately-for example, apen in one hand and a raised finger in theother-each time they are briefly moved. Thedistance between the two objects is varied.This generates reflexive saccades, which aretested in both horizontal and vertical planes,and the examiner should observe saccadicvariables such as speed of initiation (latency),accuracy, and velocity. Saccadic accuracy canbe determined by noting the size and direc-tion of any corrective saccades. Carefulobservation allows detection of saccades assmall as 0.50. Normal subjects often show aminor degree of saccadic undershoot, whichis more often found with larger gaze shifts.Any slowing of saccades can be accentu-

ated by using an optokinetic striped drum ortape, when the repositioning saccades willappear slowed. This is of particular helpwhen showing slowed adduction in a partialinternuclear ophthalmoplegia.'5 Anothermethod to accentuate this abnormality is touse oblique targets. Because the velocity isslowed in one plane the resulting saccade willbe L shaped, because the normal verticalcomponent is completed before the abnormalhorizontal one.

Other types of saccades can also be gener-ated at the bedside.'6 For example, antisac-cades (saccades in the opposite direction tothe target) can be tested by holding up bothhands on either side of the patient and askingthem not to look at the finger that is movedbut instead to look in the opposite direction.If the patient repeatedly makes a reflexive sac-cade to the finger that has moved rather thanin the opposite direction a high level of dis-tractibility (impaired ability to suppressreflexive saccades) is shown. Elevated antisac-cade error rates are associated with frontallobe dysfunction, and have been reported inpatients with frontal lobe lesions,'7 schizo-

phrenia, 18 Huntington's disease, "I progressivesupranuclear palsy, and corticobasal degener-ation,20 and motor neuron disease.7 Predictivesaccades can be tested by alternately raising afinger of one hand and then the other in apredictable regular pattern, and asking thepatient to make saccades to the target.

Finally the patient should be observed forany head movements or blinks before makinga saccade, as occurs in Huntington's disease5and ocular motor apraxia.2'

SMOOTH PURSUITSmooth pursuit can be tested by asking thepatient to track a small target at a distance ofabout 1 m while keeping their head still.Assessment of both horizontal and verticalsmooth pursuit should be performed. Thetarget should be moved initially at a slow uni-form speed and the pursuit eye movementsobserved to determine whether they aresmooth, or broken up by catch up saccades.This is a non-specific sign if present in bothdirections-for example, due to aging22 orcerebellar disease,23 or it may indicate afocal posterior cortical lesion if only presentin one direction.24 The speed should be grad-ually increased, but at high velocities allsmooth pursuit eye movements will be brokenup by saccades, even in normal subjects.These saccadic intrusions occur as thesmooth pursuit velocity of the eye fails tomatch that of the target-that is, the pursuitgain (pursuit velocity divided by target veloc-ity) is reduced.The OKN drum and tape actually stimu-

late pursuit eye movements rather than opto-kinetic nystagmus.' They are useful in testingfor pursuit asymmetries (pursuit which isworse in one direction), and reversed pursuitin which the fast phases are in the direction ofthe drum rotation, as in congenital nystag-mus.25 If the smooth pursuit gain is reducedto one side, the eyes will deviate less quicklyfrom their primary position, so fewer quickphases are seen when the subject observes anOKN drum.

If a target is tracked by head movementsthe VOR will act to generate eye movementsthat would compensate for the head move-ments. These are in an inappropriate direc-tion and are therefore normally suppressed.Studies in normal subjects26 and patients27have led to the belief that the suppression ofthe VOR is derived from the smooth pursuitsystem, and that patients with abnormalsmooth pursuit have impaired suppression ofthe VOR. This may be tested by asking thepatient to fixate their thumbnail with theirarms outstretched while rotating their headand trunk in harmony.28 Impaired cancella-tion of the VOR and hence abnormal pursuitare shown by observing the eye repeatedlymoving off fixation due to the VOR, followedby refixation saccades. This is a particularlyuseful technique for testing pursuit in patientswith gaze evoked nystagmus.

OPTOKINETIC NYSTAGMUSAs discussed earlier the optokinetic system

120 on 7 S



j.com/

J Neurol N


ugust 1995. Dow

nloaded from


Eye movements

cannot be tested as part of a clinical examina-tion because the OKN drum and tape com-

monly used actually test smooth pursuit andnot the optokinetic system.

VESTIBULAR SYSTEMThe patient should be observed for nystag-mus with and without fixation and also aftershaking the head for 15 seconds.29 Frenzelgoggles may be useful for studying eye move-

ments without the subject fixating.If the vestibulo-ocular system is function-

ing normally passive rotation of the patient'shead should result in a slow eye movement so

that the eyes deviate in the opposite directionto that of the head movement. This is knownas the doll's head (oculocephalic) manoeuvre

and should be performed both horizontallyand vertically. This technique is not onlyvaluable for assessing vestibular function, butalso for differentiating between nuclear andsupranuclear gaze palsies, and in the evalua-tion of brainstem function in comatosepatients. It should be noted that the eyemovements elicited by this procedure inunconscious patients largely reflect theintegrity of the semicircular canals and theircentral connections, although in consciouspatients the effects of visual input on eyemovements may influence the response tohead rotation.A rough estimate of any deterioration of

vestibular gain (head velocity divided by eye

velocity) can be obtained by asking thepatient to read a Snellen chart while theirhead is being passively rotated. If there is an

abnormality the visual acuity will show a

deterioration compared with the acuityobtained with the head still. With an ophthal-moscope the examiner can observe thepatient's optic disc while they fixate a distantobject and shake their head from side toside.30 If the gain of the VOR is unity theexaminer will not see any movement of thepatient's disc.

Vestibular nystagmus can be elicited byrotating the patient in a swivel chair at a con-

stant velocity. By altering head position hori-zontal, vertical and torsional nystagmus can

be shown. Postrotatory nystagmus will beseen if the chair is suddenly stopped. Thiswill be emphasised if the patient wears

Frenzel goggles, so removing fixation andallowing the examiner to estimate the dura-tion of the postrotatory nystagmus.

Caloric stimulation can also be used to testthe vestibulo-ocular system. The test is per-formed with the patient supine. The externalauditory meatus is checked to ensure that thetympanic membrane is intact. The head isflexed to 300 to allow maximum stimulationof the horizontal semicircular canals. One totwo hundred millilitres of either warm water(44°C) or cold water (30°C) is infused into

the patient's ear. The normal response in an

awake patient is for a slow phase response

towards the side of the cold water irrigation(or away from the warm water) with the fastphase away from the cold water (or towardsthe warm water). The test is best performed

both with and without fixation to assess thedegree of suppression of the nystagmus byfixation. Frenzel goggles can be used toremove fixation.

Oculographic techniquesAlthough a correct diagnosis of many eyemovement disorders can be ascertained bythe bedside methods already described, it isonly by using oculographic recording tech-niques that more subtle abnormalities can bedetected by the quantitative analysis and eval-uation they afford. In particular the develop-ment of very precise behavioural paradigms inthe laboratory evoking saccadic and pursuiteye movements has shown disturbances pre-viously undetected at the bedside.

THE IDEAL SYSTEMWhen considering the various techniquesavailable it is necessary to be clear about therequirements for an ideal eye movementrecording system.3' These are: (a) easy, non-traumatic application, ideally with no contactwith the eye. (b) No interference with normalvision, and a sufficiently large field of vision.(c) Simultaneous measurement of horizontal,vertical, and torsional eye movements. (d)High accuracy and repeatability, with a widelinear range of over 900 of eye position. (e)High resolution allowing detection of eyemovements as small as a few seconds of arc.(f) Good stability with no baseline drift. (g)Good dynamic measuring range (frequencybandwidth) of zero to a few hundred Hertz.(h) Insensitivity to translational head move-ments, and thus no need for rigid head fixa-tion. (zT) Insensitivity to surrounding levels ofillumination, and to artefacts arising fromblinks and electromyographic or electro-mechanical interference.

Unfortunately none of the current tech-niques fulfil all these criteria, although someof the four main methodologies in current usefulfil most of them.

ELECTRO-OCULOGRAPHYThe most commonly used technique and theone that has been available for the longestperiod is electro-oculography (EOG). As amethod to provide a simple visual record ofan eye movement disorder EOG is the sim-plest and cheapest. If high resolution analysisof eye movements, particularly saccades isrequired, however, this technique has numer-ous disadvantages. It relies on the fact thatthe globe acts as an electrostatic dipole, withthe cornea 0-4-1 0 mV positive with respectto the posterior pole. Electrodes are placedon the skin on either side of the eye and themagnitude of the potential recorded dependson the proximity of the cornea to the elec-trode.32 A major disadvantage is the variabilityof the corneoretinal potential, particularly inrelation to ambient illumination.33 Thisresults in drifting of the baseline, and it is,therefore, necessary to dark adapt the subjectfor 15-20 minutes before testing. There isalso a rather high baseline noise level due

121 on 7 S



j.com/

J Neurol N


ugust 1995. Dow

nloaded from



both to electromagnetic interference and peri-orbital muscle activity.34 This necessitates theuse of low pass filtering with a differentialamplifier and great care with the placement ofthe electrodes. Its sensitivity is at best 0.50and more usually only 1-2°.

There are advantages to EOG over othertechniques. There is a large measurementrange of up to ±60° horizontally and ±400vertically, although the vertical movementrecordings are subject to error and only pro-

vide qualitative assessment; the analogue out-put is linear over a range of ±30° 34; there are

no limitations of the fields of vision and thesubjects can wear their spectacles; andrecordings can be made with the eyelidsclosed, which is useful during vestibular andcaloric testing.

INFRARED LIMBUS REFLECTION OCULOGRAPHYThe second technique that is being increas-ingly used both for routine eye movementassessment and research is the infrared lim-bus reflection (IRLR) technique. The eye isdiffusely illuminated with infrared, which isdifferentially reflected by the iris and thesclera back to photocells positioned near thelimbus.35 As the sclera reflects better than theiris the output from the photocells will bedirectly proportional to the position of theeye. The positioning of the infrared emittersand the photocells is crucial for accurate andlinear recordings.36 The other problem is fixa-tion of the head, which is necessary if therecorded eye movement is to be an accuraterepresentation of direction of gaze. Infraredoculography systems are generally linearwithin a horizontal range of ± 150, and a ver-

tical range of ± 100, although satisfactory ver-

tical recordings can be difficult.37 It has beenclaimed that sensitivity for this technique is ofthe order of 3 minutes of arc, but in practiceit is not quite so impressive.

HIGH SPEED VIDEO RECORDERSA third technique is high speed televisionrecording, which with the rapid advances intechnology, and therefore of temporal resolu-tion, is likely to become increasingly used foroculography. These systems usually locate thepupil centre using software algorithms foreach frame.38 Currently the spatial resolutionis very good in both horizontal and verticalaxes, and is highly suitable for recording two

dimensional scanpaths, but temporal resolu-tion is still inadequate for full analysis of sac-

cadic metrics.

MAGNETIC FIELD SCLERAL SEARCH COIL

TECHNIQUEFinally, probably the most accurate system isthe magnetic field scleral search coil(MFSSC) method.39 The subject is seated so

that their head is located in an alternatingmagnetic field. The subject wears a ringshaped contact lens in which is embedded a

coil of wire. An alternating current is therebyinduced in the wire, which is proportional to

the sine of the angle between the planes ofthe search coil and the direction of the mag-

netic field. Both horizontal and vertical eyeposition may be measured simultaneously,and with appropriate windings of the wire coilso may torsional eye movements. Themethod has low noise levels and high sensitiv-ity in the order of a few seconds of arc,3940and its linear range is ±200, greater if a sinefunction correction is used.

There are, however, two main disadvan-tages of the method. Firstly, the subject hasto wear a scleral contact lens held to the eyeby suction.41 Although topical local anaes-thetic is used, some subjects find the insertionand removal of the coil unpleasant. As thecoil can only be left in place for up to 30 min-utes this can limit the number of tests per-formed. The second main disadvantage iscost. Most complete systems are expensive asare the coils, which usually last for only aboutfive subjects.

It becomes apparent that none of the cur-rently available techniques meets the full setof ideal criteria discussed earlier. The methodchosen rather depends on the clinician'srequirements. If, for example, a record of apatient's nystagmus is required, then astraightforward video recording would beappropriate, perhaps with an EOG chartrecord. For recording slow eye movementssuch as VOR or OKN the use of EOG is ade-quate, but if a detailed analysis of saccadicmetrics is required IRLR or MFSSC wouldbe more appropriate. The cheapest method isEOG and MFSSC is the most expensive. Iflarge eccentric eye movements are to berecorded then EOG is best, as it is for record-ing patients with their eyes closed, required intesting vestibular function, or when uncon-scious.

Although it is possible to perform somequantification of eye movement recordsdirectly from the chart records, the analoguesignal produced by the oculographic equip-ment is usually digitised and the data analysisproceeds off line at a later date with interac-tive computer programs.

Laboratory investigation of dynamic eyemovementsSACCADESEssentially when saccadic eye movements areinvestigated in the laboratory, the same para-digms described earlier are tested. A range ofmore behaviourally demanding saccadic para-digms such as the antisaccade, predictive sac-cade, and remembered saccade tasks may beused in the laboratory, and may show a disso-ciation in performance in different patientgroups.16 The measurement of saccades in thelaboratory also allows both more controlledpresentation of the stimuli and more accuratemeasurement of the variables mentioned. Thebeginning and end of a saccade are definedarbitrarily, often when the eye velocity risesabove or falls below 200/s respectively.

Rapid refixation gaze movements arelargely achieved by a saccade that coversabout 90% of the required distance (primarysaccade), followed by a series of secondary

122 on 7 S



j.com/

J Neurol N


ugust 1995. Dow

nloaded from


Eye movements

Figure 2 Recordings ofhorizontal eye movementstaken from two subjectsperforming the antisaccadetask (magnetic field scleralsearch coil technique).Upward displacementsindicate movement of theeye or target to the rightand downwarddisplacement movements tothe left. (A) Controlsubject; the subject isperforming the paradigmcorrectly, and makingsaccades in the oppositedirection to the target. (B)Subject with frontal lobedysfunction. The arrowsdenote the onset of reflexivesaccades in the samedirection as the target,which are corrected in allcases by saccades in theopposite direction.

A

15Target I degrees

Eye X l X XI degrees

B

Target degrees

4' + 4 ; 4

Eye | degrees

2500ms

saccades until the target is foveated. Accuracyof the primary saccade is expressed as sac-cadic gain. This is the value of the primarysaccadic amplitude divided by the targetamplitude, so that a perfectly accurate pri-mary saccade has a gain of 1 0. Often the gainof the primary saccade and also the gain ofthe final eye position (FEP) after the sec-ondary saccades are recorded. Figure 2 showsan example of a laboratory recording of per-formance in a saccadic task.The peak velocities of saccades show a

unique feature; there is a progressive expo-nential increase in velocity with saccadic

amplitude until it saturates at about 20-30°.This consistent relation between saccadicamplitude and peak velocity is known as themain sequence (fig 3).42 Clinical laboratoriesshould derive their own normative values forsaccadic variables, as these vary depending onthe precise details of the test procedures andthe oculographic equipment.

SMOOTH PURSUIT (FIG 4)Laboratory evaluation of smooth pursuit usesmany different types of stimulus presentation,some of which cannot be tested at the bed-side. Usually the stimulus is a small brightspot of light, the position of which is con-trolled by mirror galvanometers, and which isprojected on to a screen. Alternatively a spoton a computer screen may be used althoughthis method does not allow testing of such alarge range of movements.The most commonly used pursuit stimuli

are either constant velocity (ramp stimulus)or sinusoidal waveforms. A range of differentvelocities or frequencies are tested.Interactive computer programs allow theanalysis of smooth pursuit to calculate suchpursuit variables as velocity gain (eye veloc-ity/target velocity). The maximum gain aver-aged from a series of cycles or the gainaveraged as the eye passes the primary posi-tion is measured. For sinusoidal targets bothgain and phase can be measured. The gain iscalculated from peak eye velocity/peak targetvelocity. The gain is known to be dependenton the peak acceleration of the target.43 Thesestimuli mainly test the maintenance of pur-suit, although another aspect of the pursuitsystem of interest is pursuit initiation.44

OPTOKINETIC NYSTAGMUSAdequate assessment of the optokinetic sys-tem requires a stimulus that fills the patient's

Figure 3 Main sequence:a graph of the peakvelocity ofsaccades from11 normal subjects plottedas a function of theiramplitude.

(ACo0)01)03)0L)

0

0)a)

a)

0-

500 r

450 H

400 H

350 H

300 F

250 H

* . . .

;040 1

*.': ..0 ..

a.8

200 H

150 H

100 H

50 H

011 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15

Saccade amplitude (degrees)

123

:,0

on 7 Septem

ber 2018 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.59.2.115 on 1 August 1995. D

ownloaded from



Figure 4 Oculographicrecordings of smoothpursuitfrom two subjectstracking a target movinghorizontally at a constantvelocity of 200 (magneticfield scleral search coiltechnique). (A) Tracingfrom a normal subjectshowing accurate tracking.(B) Tracingfrom a patientwith motor neuron diseaseshowing poor tracking withfrequent catch up saccades.

A B

Target

Eye

22g5degrees

T22.5Idegrees

Velocity I 1501degrees/s

2000 ms

field of vision and induces the sense of selfrotation (circularvection). One method is torotate the patient at a constant velocity in thelight for more than a minute. This allows theVOR to decay and so the nystagmus is solelydue to visual input. Another method is to sitthe patient inside a large optokinetic drum.The eye movements stimulated are conjugateeye movements of constant velocity in thedirection of the stimulus, interspersed withquick phases in the opposite direction. Theway in which the patient is instructed deter-mines the type of response. If the patient isasked to follow the stripes look nystagmus,which consists of prolonged slow phases withlarge corrective saccades, is seen. If the sub-ject is asked to stare ahead at the stripes starenystagmus is induced, which is characterisedby smaller and more frequent quick phases.

Both the smooth pursuit system and theoptokinetic system contribute to theresponse. During the initial few secondssmooth pursuit constitutes the more impor-tant component. When the visual stimulus

Table S Clinical and investigative approach to eye movement abnormalities

Bedside assessment

Ocular ductions and versionsHirschberg/Krimsky testCover-uncover testAlternate cover testParks-Bielschowsky test

Laboratory investigation

Static eye movements (diplopia)Hess screen testRed glass testLancaster red-green testMaddox rod testTensilon testForced ductions

Dynamic eye movements

Fixation stabilityLatency, accuracy, velocityDistractibility (for example, antisaccades)

Smooth pursuit:

Saccadic intrusionsSymmetry/asymmetryVestibulo-ocular reflex suppression

Vestibulo-ocular reflex:Doll's head manoeuvreCaloric stimulation

Optokinetic nystagmus:OKN drum(actually tests smooth pursuit; see text)

EOG, IRLR, MFSSC, highspeed video

Variables studied:Primary saccade gainFEP gainLatency, velocityDistractibility

EOG, IRLR, MFSSC, highspeed video

Variables studied:Velocity gainsaccadic intrusionsPhase

See Balance review

Full field visual stimulation

ceases the response should continue. This isknown as optokinetic afternystagmus(OKAN). The optokinetic system acts as avelocity storage mechanism during the pres-ence of the stimuli and when this is no longervisible the nystagmus continues in the samedirection for a few seconds with a decliningslow phase velocity.When studying OKN the nystagmus seen

in the presence and absence of visual stimulishould be studied. This is most easily done byturning out the lights after a period of optoki-netic stimulation, often after the drum hasrotated at 60°/s for 60 seconds. The velocitygain (eye velocity/target velocity) and thesymmetry of the response are measured forboth OKN and OKAN. The time constant ofthe slow phase eye movements of the OKANcan also be measured.

Because of the great variability in the mea-surement of a subject's OKAN eye velocityand time constant, repeated measures areneeded.45 Alternatively the build up of theslow phase velocity of the OKAN can bemonitored by briefly turning out the lightsduring frequent, two second periods of dark-ness during the stimulation.46 To preventcontamination of the data by smooth pursuitthe first one second of each two secondperiod should be discarded.

ConclusionThe intention of this review has been to pro-vide the clinician with an overview of the clin-ical assessment and appropriate investigationof a patient presenting with a disturbance ofeye movements (table 5). Because many neu-rological conditions may cause abnormalitiesof ocular motility, a rational approach to thissituation remains of great importance to thepractising neurologist. Although a variety ofspecialised investigations are available, anappreciation of the basic anatomical andphysiological principles involved is necessary,both for an adequate bedside assessment andfor such investigations to be appropriatelyemployed and interpreted. These findingsmust of course be viewed in conjunction with

Saccades:

124 on 7 S



j.com/

J Neurol N


ugust 1995. Dow

nloaded from


Eye movements

the general neurological examination and theresults of other appropriate investigations,such as neuroimaging, discussion of which isbeyond the remit of this review.

1 Leigh RJ, Zee DS. The neurology of eye movements. 2nd ed.Philadelphia: FA Davis, 1991.

2 Sharpe JA, Herishanu YO, White OB. Cerebral squarewave jerks. Neurology 1982;32:57-62.

3 Rascol 0, Sabatini U, Simonetta-Moreau M, et al. Squarewave jerks in Parkinsonian syndromes. J NeurolNeurosurg Psychiatry 1991 ;54:599-602.

4 Elidan J, Gay I, Lev S. Square wave jerks-incidence, char-acteristics and significance. J Otolaryngol 1984;13:375-81.

5 Leigh R, Newman S, Folstein S, et al. Abnormal ocularmotor control in Huntington's chorea. Neurology 1983;33:1268-75.

6 Ross D, Ochs A, Hill M, et al. Erratic eye tracking inschizophrenic patients as revealed by high-resolutiontechniques. Biol Psychiatry 1988;24:675-88.

7 Shaunak S, Orrell R, O'Sullivan E, et al. Oculomotorfunction in amyotrophic lateral sclerosis: evidence forfrontal impairment. Ann Neurol 1995 (in press).

8 Smith JL. Monocular diplopia [editorial comment].J7ournal of Clinical Neuroophthalmology 1986;6: 184-5.

9 Mein J, Trimble R. Diagnosis and management of ocularmotility disorders. 2nd edition. Oxford: Blackwell, 1991.

10 Lancaster WB. Detecting, measuring, plotting and inter-preting ocular deviations. Arch Ophthalmol 1939;22:867-80.

11 de Waard R, Koorneef L, Verbeeten G. Motility distur-bances in Graves ophthalmopathy. Doc Ophthalmol1983;56:41-7.

12 Moore AT, Morin JD. Bilateral acquired inflammatoryBrown's syndrome. J Pediatr Ophthalmol Strabismus1985;22:26-30.

13 Zappia RJ, Winkelman JZ, Gay AJ. Intraocular pressurechanges in normal subjects and the adhesive musclesyndrome. Am J Ophthalmol 197 1;71:880-3.

14 Moorthy G, Behrens M, Drachman D, et al. Ocularpseudomyasthenia or ocular myasthenia "plus": a warn-ing to clinicians. Neurology 1989;39:1150-4.

15 Smith JL, David NJ. Internuclear ophthalmoplegia. Twonew clinical signs. Neurology 1964;14:307-9.

16 Kennard C, Crawford TJ, Henderson L. A pathophysio-logical approach to saccadic eye movements in neuro-logical and psychiatric disease. 7 Neurol NeurosurgPsychiatry 1994;57:881-5.

17 Guitton D, Buchtal H, Douglas R. Frontal lobe lesions inman cause difficulties in suppressing reflexive glancesand in generating goal directed saccades. Exp Brain Res1985;58:455-74.

18 Fukushima J, Fukushima K, Chiba T, et al. Disturbancesof voluntary control of saccadic eye movements in schiz-ophrenic patients. Biol Psychiatry 1988;23:670-7.

19 Lasker A, Zee D, Hain T, et al. Saccades in Huntington'sdisease: Initiation defects and distractibility. Neurology1987;37:364-70.

20 Vidailhet M, Rivaud S, Gouidet-Khouja N, et al. Eyemovements in Parkinsonian syndromes. Ann Neurol1994;35:420-6.

21 Zee DS, Yee RD, Singer HS. Congenital ocular motorapraxia. Brain 1977;100:581-99.

22 Zackon D, Sharpe J. Smooth pursuit in senescence. ActaOtolaryngol (Stockh) 1987;104:297-8.

23 Pierrot-Deseilligny C, Amarenco P, Roullet E, MarteauR. Vermal infarct with pursuit eye movement disorders.J Neurol Neurosurg Psychiatry 1990;53:519-21.

24 Morrow MJ, Sharpe JA. Cerebral hemispheric localisationof smooth pursuit asymmetry. Neurology 1990;40:284-92.

25 Halmagyi GM, Gresty MA, Leech J. Reversed optokineticnystagmus (OKN): mechanism and clinical significance.Ann Neurol 1979;6:80-3.

26 Barnes GR, Benson AJ, Prior ARJ. Visual-vestibular inter-action in the control of eye movement. Aviat SpaceEnviron Med 1978;49:557-64.

27 Halmagyi GM, Gresty MA. Clinical signs of visual-vestibular interaction. J Neurol Neurosurg Psychiatry1979;42:934-9.

28 Dichgans J, Von Reutern GM, Rommelt U. Impaired sup-pression of vestibular nystagmus by fixation in cerebellarand noncerebellar patients. Archiv Afir Psychiatnie undNervenkrankheiten 1978;226: 183-99.

29 Hai TC, Fetter M, Zee DS. Head shaking nystagmus inpatients with unilateral peripheral vestibular lesions. AmJ Otolaryngol 1987;8:36-47.

30 Zee DS. Ophthalmoloscopy in examination of patientswith vestibular disorders. Ann Neurol 1978;3:373-4.

31 Lueck CJ, Kennard C. Oculography and techniques foreye movement recording. Bull Soc Belge Ophthalmol1989;237:485-502.

32 Shackel B. Eye movement recording by electrooculogra-phy. In: A manual of psychophysiological methods.Venables PH, Martion I, eds. Amsterdam: NorthHolland Publishing, 1967:300-34.

33 Arden FH, Kelsey JH. Changes produced by light in thestanding potential of the human eye. J7 Physiol 1962;161:189-204.

34 Young LR, Sheena D. Survey of eye movement recordingmethods. Behaviour Research Methods and Instrumen-tation 1975;7:397-429.

35 Stark L, Vossius G, Young LR. Predictive control of eyetracking movements. Institute of Radio EngineeringTransactions on Human Factors in Electronics 1962;3:52-7.

36 Truong DM, Feldon SE. Sources of artefact in infraredrecording of eye movements. Invest Ophthalmol Vis Sci1987;28:1018-22.

37 Gauthier GM, Volle M. Two-dimensional eye movementmonitor for clinical and laboratory recordings.Electroencephalogr Clin Neurophysiol 1975;39:285-9 1.

38 Barbur JL, Thomson WD, Forsyth PM. A new system forthe simultaneous measurement of pupil size and three-dimensional eye movements. Clin Vis Sci 1987;2:131-42.

39 Robinson DA. A method of measuring eye movementusing a scleral coil in a magnetic field. IEEE TransactionsofBiomedical Electronics 1963;10: 137-45.

40 Remmel RS. An inexpensive eye movement monitor usingthe scleral search coil technique. IEEE Tran Biomed Eng1984;31 :388-90.

41 Collewijn H, Van der Mark F, Jansen TC. Precise record-ing of human eye movements. Vision Res 1975;15:447-50.

42 Boghen D, Troost BT, Daroff RB, et al. Velocity charac-teristics of normal human saccades. Invest OphthalmolVis Sci 1974;13:619-23.

43 Lisberger SG, Evinger C, Johanson GW, Fuchs AF.Relationship between eye acceleration and retinal imagevelocity during foveal smooth pursuit in man and mon-key. J Neurophysiol 198 1;46:229-49.

44 Thurston SE, Leigh RJ, Crawford TJ, et al. Two distinctdeficits of vision tracking caused by unilateral lesions ofcerebral cortex in humans. Ann Neurol 1988;23:266-73.

45 Tijssen MAJ, Straathof CSM, Hain TC, Zee DS.Optokinetic afternystagmus in humans: normal valuesof amplitude, time constant and asymmetry. Ann OtolRhinol Laryngol 1989;98:741-6.

46 Segal BN, Liben S. Modulation of human velocity storagesampled during intermittently-illuminated optokineticstimulation. Exp Brain Res 1985;59:515-23.

125 on 7 S



j.com/

J Neurol N


ugust 1995. Dow

nloaded from


eye movements - jnnp.bmj.com · magneticfield scleral search coil oculographic technique required...

Documents