3rd Congress of the European Academy of Neurology

Amsterdam, The Netherlands, June 24 – 27, 2017

Hands-on Course 3

Bedside examination of the vestibular and ocular motor system - Level 2

How to examine the vestibular system

Raymond van de Berg Maastricht, The Netherlands

Email: raymond.vande.berg@mumc.nl

1

Conflict of interest: The author has no conflict of interest in relation to this manuscript

Introduction

A complete and thorough neuro-otological and vestibular examination

is necessary to find any signs of vestibular hypofunction or any neuro-

logical diseases, particularly ataxia. Keep in mind that when testing the

vestibular system, each test has its own limitations in terms of sensitivity,

specificity, patient acceptance, costs and duration. During the neuro-

otological assessment, one should pay especially close attention to the

oculomotor examination, since abnormal oculomotor findings may be the

only or first presenting central signs that may explain the vestibular

symptoms [2]. The oculomotor examination is best performed before

inducing the substantial head movements that are typical for some major

components of the vestibular examination. The vestibular examination

includes the Dix-Hallpike and the lateral roll test, positional testing,

(V)HIT, the test for Dynamic Visual Acuity (DVA), the visually enhanced

vestibulo-ocular reflex test, fixation suppression, the Valsalva maneuver

(straining against the closed glottis and blowing out against pinched

nostrils), the head shake test, the vibration test, the hyperventilation test

and the Romberg test on foam rubber or in tandem [3, 4]. The Romberg,

(V)HIT, DVA and head shake test will be discussed more in detail in this

hands-on course.

2

The Romberg test

The Romberg test mainly diagnoses ataxia and is not specific for a

vestibular loss, since it also detects cerebellar and proprioceptive impair-

ment [3, 5]. However, the sensitivity for detecting vestibular deficits

increases when the patient stands on foam rubber [6]. The Romberg test

on foam rubber has a sensitivity of up to 79% and a specificity of up to 80%

for detecting both patients with unilateral and those with bilateral

vestibular loss [3, 7].

Head impulse testing

A brief, high-acceleration head ‘impulse’ can test vestibular function of

all semicircular canals. Depending on the semicircular canal tested, the

head is rotated in a different direction [8, 9]. A corrective catch-up

saccade is made in case of vestibular hypofunction. HIT can be performed

with or without the use of a noninvasive video-oculography device (i.e.

Video Head Impulse Test: VHIT). This device consists of goggles that

contain a high-speed infrared video camera that tracks eye movements

and accelerometers that track head movements [10].

Although applying HIT may sound simple at first, some challenges are met

when performing it. The first challenge is to adequately deliver the

stimulus: it should be a high acceleration (1,000–6,000°/s2), rapid (100–

200°/s), low-amplitude (10–20°) head rotation. The horizontal as well as

the vertical canals can be tested. For the horizontal canals, the head is

moved in the horizontal plane. This is different for the vertical canals. For

this, it is important to know that the vertical canals form two pairs: The

left anterior and the right posterior canal (they are called LARP) and the

right anterior and left posterior canal (they are called RALP). They form

3

pairs, since the anterior canal of one side, is oriented in the same plane as

the posterior canal of the other side. This can be seen in Figure 1 [11].

Figure 1. Orientation of the canals

This means that a head movement in the plane of for example LARP, has

effect on both canals. The anterior canal is sensitive for rotations going

down in its plane, and the posterior canal is sensitive for rotations going

up in its plane. Therefore, if you want to test the left anterior canal for

example, the following steps have to be taken: First bring the head in line

of the canal, so turn the head to the right. Secondly, have the patient

fixate on a target, for example your nose. Thirdly, make an impulse in line

of the canal and use the direction in which the canal is most sensitive.

Since the anterior canal is most sensitive for moving down, make a down-

wards movement. The same paradigm can be used for the other canals.

This is shown in Figure 2a and 2b [11].

Figure 2a. Testing LARP Figure 2b. Testing RALP

4

For both horizontal and vertical impulses, the head movements can be

applied inward or outwards (figure 3). With the inward impulses, the head

is first displaced laterally and then rapidly rotated back to the center.

With the outward impulses, the head is put in neutral position and then

rapidly rotated toward either side. The inward impulses might be easier

to interpret in patients with acute unilateral peripheral vestibulopathy,

where spontaneous nystagmus will vary in intensity depending on orbital

position (figure 3). Also, less bounce artifacts are present. However,

outward impulses are less predictable, since the neck offers no clue as to

the direction of the next impulse. This might reduce the risk of covert

saccades (see below). Therefore, the risk of erroneously identifying a

normal HIT is higher using the inward technique [12].

Figure 3. Two head impulse techniques in a patients with left unilateral

vestibular deficit: inward HITs toward the deficient labyrinth with an

eccentric head and eye position before the HIT, and outward HITs to the

left, where the head and eye start moving from a centered position. The

final resting head and eye position, which is eccentric in outward HITs, is

important. Such an eccentric eye position would enhance an underlying

spontaneous nystagmus according to Alexander’s law and, thus,

potentially confound test interpretation [12].

5

When using VHIT, one should pay attention by avoiding a loose strap,

wrong calibration, pupil tracking loss, (mini-)blinks, touching the goggles,

patient inattention and investigator induced bounce; if these are not

avoided, they will result in artifacts [13].

The second challenge is not to be fooled by preprogrammed compensatory

saccades (‘covert saccades’) that can be invisible to the naked eye of the

examiner and can occur (not only) in bilateral vestibulopathy patients.

Consequently, vestibulopathy may be missed [14]. A recent study by

Strupp et al. indicated that HIT observed by the naked eye of experts is

false negative for about 50% of the patients when compared to VHIT [15].

This clearly supports the use of the VHIT device, which is able to track

these saccades. Examples of normal and abnormal VHIT recordings with

overt and covert saccades are presented in Figure 4.1a–c.

The third challenge is to correctly interpret the traces. VHIT traces can

have many artifacts, leading up to 42% of uninterpretable traces [13].

Besides these artifacts, eye movements in patients with a vestibular

hypofunction can show patterns that challenge interpretations. Ideally,

vestibulo-ocular reflex gain is calculated by peak eye velocity divided by

peak head velocity [16]. However, artifacts and abnormal patterns distort

the process of correct gain calculation, and commercially available soft-

ware is not yet able to adequately compensate for it. An example of an

eye movement pattern that interferes with gain calculation is presented

in Figure 4.2. In order not to miss a bilateral vestibulopathy, a physician

should not yet solely rely on software processing for gain calculation, but

should be trained in assessing the raw data and should be aware of the im-

pact of deviant eye movement patterns and measurement artefacts [13].

6

Figure 4.1 Raw VHIT recordings of different subjects, recorded with the

EyeSeeCam system (EyeSeeCam VOG; EyeSeeCam, Munich, Germany).

Head velocity traces are shown in gray, eye velocity traces in black. a VHIT recordings of head impulses to the right in a healthy subject. The

eye movements compensate for the passive head movements. b VHIT

recordings of head impulses to the left in a patient with a peripheral

vestibular deficit, resulting in overt saccades (peaks in eye velocity after

head movements). The eye movements do not compensate for the passive

head movements. c VHIT recordings of head impulses to the right in a

patient with a peripheral vestibular deficit, mainly resulting in covert

saccades (peaks in eye velocity during head movements).

300Rightn= 3

200EyeHead

100

0 –100 –200

–3000 100 200 300

c Time(ms)

Velocity (°/s)

300Leftn= 4

200EyeHead

100

0 –100 –200

–3000 100 200 300

b Time(ms)

Velocity (°/s)

300Rightn= 4

200EyeHead

100

0 –100 –200

–3000 100 200 300

a Time(ms)

Velocity (°/s)

7

Figure 4.2 Example of an eye movement pattern that interferes with gain

calculation. Raw VHIT recordings of head impulses to the left in a patient

with bilateral vestibulopathy are presented. Head velocity traces are

shown in gray, eye velocity traces in black. Normally, gain is calculated by

peak eye velocity divided by peak head velocity. In this case, gain

calculation is challenged, since at the moment the peak head velocities

are reached, the eyes are actually moving along with the head in the same

direction. The passive head movements are not compensated by the eye

movements. Although this is clearly an abnormal HIT, there is not yet any

consensus about how to determine (or whether it is even possible to

determine) the real peak eye velocity which is necessary for gain

calculation.

The fourth challenge is to correctly interpret the end result. HIT provides

a stimulus for measuring gain of the vestibulo-ocular reflex which is

different from those used in other vestibular tests such as the rotatory

chair tests or the caloric test; it includes many more high frequency

components than the rotatory chair tests and the caloric test. Differences

in response to the caloric test versus the rotation tests versus HIT are

especially pointing to this difference in frequency content. It has been

shown that a bilateral vestibular loss can be measured with the caloric

test, while the responses as measured with HIT are relatively preserved

[17, 18]. In other words, it is necessary to understand that the presence of

a normal vestibulo-ocular reflex as measured with HIT does not rule out a

vestibular deficiency.

300Leftn= 3

200EyeHead

100

0 –100 –200

–3000 100 200 300

Time(ms)

Velocity (°/s)

8

Dynamic visual acuity

During head movements, efficient stabilization of the image on the retina

is necessary to preserve visual acuity [19]. In (bilateral) vestibulopathy

patients, gaze stabilization fails and can lead to significant deterioration

in visual acuity during head movements [20, 21]. Visual acuity in dynamic

conditions can be assessed by testing for DVA. DVA testing can be

performed in many ways: the patient has to read letters from a visual

acuity chart or a computer screen during active or passive, vertical or

horizontal head movements, or while walking on a treadmill at different

velocities [19, 22]. Passive high-angular-velocity movements (150°/s) have

been shown to be most useful for discrimination between healthy subjects

and patients with a unilateral or bilateral vestibular loss. However, that

study did not include DVA testing by walking on a treadmill [23]. A decline

of more than 2 lines on the optotype chart is considered abnormal [24],

although a loss of 2 lines (0.2 logMAR) is not unusual for healthy subjects.

In order to trade sensitivity for specificity, 4 lines may be required [25].

Moreover, DVA may show false negative results due to mechanisms that

at least partially compensate for the retinal instability during head

movements [3, 23]. However, in subjects with unilateral and bilateral

vestibular loss, computerized DVA testing reached a sensitivity of 94.5%

and a specificity of 95.2% [26]. In another group of bilateral vestibulo-

pathy patients, DVA was impaired in 96% of the cases [27]. To conclude,

DVA can help establishing the diagnosis of (bilateral) vestibulopathy, but a

normal DVA does not definitely rule out bilateral vestibulopathy, and an

impaired DVA does not imply vestibulopathy per se. It is still not under-

stood by which specific vestibular deficits (which semicircular canals,

which otolith organs and which frequencies) DVA decreases.

9

Head shake test [4]

The head shake test is performed by shaking the head of the patient when

the patient is wearing Frenzel or video goggles. The head is rotated at a

comfortable range at a frequency of about 2 to 3 Hertz for 10- to 15

seconds. After shaking, the head remains in the center and the eyes of the

patient are observed for nystagmus. The head shake test is used to look

for signs of dynamic imbalance of vestibular function by evaluating the

post-shaking nystagmus. This test is part of the whole diagnostic test

battery and will not give enough information on its own.

Different types of nystagmus can be found when applying the head shake

test. First a schematic representation of the process of head shaking will

be used to explain the physiology: By shaking the head, the velocity

storage within the vestibular nuclei is charged. So when the head is turned

to the right a VOR is produced moving the eyes in the opposite direction.

At the same time the velocity storage is charged with each head turn to

the right. For schematic purposes, we will call this “the velocity storage

on the right”. When the head is turned to the left, the velocity storage on

the left is charged. However, in case of a vestibular hypofunction or

imbalance, the velocity storage on the affected side cannot be charged as

much as the velocity storage on the other side due to the hypofunction.

This will become apparent when suddenly the head is stopped at the end

of the shaking. At that moment, both velocity storages discharge. If one

velocity storage was better charged than the other one, this will discharge

more strongly. At that moment it gives the same result as if the head was

turning to the side of the strongest velocity storage. This results in a

nystagmus away from the affected side. Therefore, in for example an

recent vestibular loss, a post-head shaking nystagmus might be found

10

pointing towards the healthy side. In some cases, the nystagmus may

reverse direction after the initial phase. This probably reflects an adapta-

tion mechanism. However, it is not always this easy, since nystagmus does

not always point to the affected side. In some cases, the brain already

compensated to the hypofunction. This means, from a schematic point of

view, that the velocity storage on the side of the hypofunction, is equally

charged compared to the healthy side when the head is rotated. That is

why, after a longer standing vestibular loss, the head shake test does

often not show any post shaking nystagmus. However, if in the meantime

the vestibular function on the paretic side was restored, the new level of

velocity storage on the paretic side velocity storage becomes excessive

relative to the compensation. During head shaking, the velocity storage on

the paretic side becomes more charged than on the healthy side. At the

end of head shaking, the velocity storage of the paretic side will therefore

discharge stronger than the healthy side. Since nystagmus points to the

strongest side, the post shaking nystagmus will point to the paretic or

former paretic side. This is called a “recovery” nystagmus.

In some central cases, mainly of the cerebellum and brainstem, head

shaking induced nystagmus can also be found. The nystagmus may beat

ipsilaterally, but also a cross-coupled “perverted” nystagmus can be

present. This means that horizontal head shaking induces a vertical post

head shaking nystagmus. From a schematic point of view: horizontal

vestibular input is calculated wrongly by the central system and this

results in a vertical nystagmus. So when a vertical nystagmus results after

the head shaking test, always take a central vestibular disorder into

account.

11

Keypoints

Each vestibular test has its own sensitivity, specificity and challenges

when applying it in clinical practice:

- The Romberg test is not a sensitive test for vestibular hypofunction.

The Romberg test on foam rubber can increase sensitivity and

specificity.

- An abnormal clinical head impulse test is a sign of vestibular hypo-

function, but a normal clinical head impulse test does not rule out

vestibular hypofunction, mainly due to adaptation mechanisms. A

video head impulse test can help detecting these adaptive

mechanisms.

- The Dynamic Visual Acuity test can help establishing the diagnosis of

(bilateral) vestibulopathy, but a normal DVA does not definitely rule

out bilateral vestibulopathy, and an impaired DVA does not imply

vestibulopathy per se.

- The head shake test is used to look for signs of dynamic imbalance of

vestibular function by evaluating the post-shaking nystagmus (slow

phases directed toward affected ear, or recovery nystagmus, or

perverted nystagmus). This test is part of the whole diagnostic test

battery and will not give enough information on its own.

12

References

1. van de Berg, R., M. van Tilburg, and H. Kingma, Bilateral

Vestibular Hypofunction: Challenges in Establishing the Diagnosis

in Adults. ORL J Otorhinolaryngol Relat Spec, 2015. 77(4):

p. 197-218.

2. Kattah, J.C., et al., HINTS to diagnose stroke in the acute

vestibular syndrome: three-step bedside oculomotor examination

more sensitive than early MRI diffusion-weighted imaging. Stroke,

2009. 40(11): p. 3504-10.

3. Petersen, J.A., D. Straumann, and K.P. Weber, Clinical diagnosis of

bilateral vestibular loss: three simple bedside tests. Ther Adv

Neurol Disord, 2013. 6(1): p. 41-5.

4. Kheradmand, A. and D.S. Zee, The bedside examination of the

vestibulo-ocular reflex (VOR): an update. Rev Neurol (Paris), 2012.

168(10): p. 710-9.

5. Khasnis, A. and R.M. Gokula, Romberg's test. J Postgrad Med, 2003.

49(2): p. 169-72.

6. Lanska, D.J. and C.G. Goetz, Romberg's sign: development,

adoption, and adaptation in the 19th century. Neurology,

2000. 55(8): p. 1201-6.

7. Fujimoto, C., et al., Assessment of diagnostic accuracy of foam

posturography for peripheral vestibular disorders: analysis of

parameters related to visual and somatosensory dependence.

Clin Neurophysiol, 2009. 120(7): p. 1408-14.

8. Halmagyi, G.M. and I.S. Curthoys, A clinical sign of canal paresis.

Arch Neurol, 1988. 45(7): p. 737-9.

9. Migliaccio, A.A. and P.D. Cremer, The 2D modified head impulse

test: a 2D technique for measuring function in all six semi-circular

canals. J Vestib Res, 2011. 21(4): p. 227-34.

10. MacDougall, H.G., et al., The video head impulse test: diagnostic

accuracy in peripheral vestibulopathy. Neurology, 2009. 73(14):

p. 1134-41.

11. Bronstein, A.M., Oxford textbook of vertigo and imbalance. 2013,

Oxford, U.K.: Oxford University Press. xii, 354 p.

12. Mantokoudis, G., et al., Compensatory saccades benefit from

prediction during head impulse testing in early recovery from

vestibular deafferentation. Eur Arch Otorhinolaryngol, 2016.

273(6): p. 1379-85.

13

13. Mantokoudis, G., et al., Quantifying the vestibulo-ocular reflex

with video-oculography: nature and frequency of artifacts. Audiol

Neurootol, 2015. 20(1): p. 39-50.

14. Weber, K.P., et al., Horizontal head impulse test detects

gentamicin vestibulotoxicity. Neurology, 2009. 72(16): p. 1417-24.

15. Yip, C.W., et al., Comparison of the Bedside Head-Impulse Test

with the Video Head-Impulse Test in a Clinical Practice Setting: A

Prospective Study of 500 Outpatients. Front Neurol, 2016. 7: p. 58.

16. Weber, K.P., et al., Head impulse test in unilateral vestibular loss:

vestibulo-ocular reflex and catch-up saccades. Neurology,

2008. 70(6): p. 454-63.

17. Agrawal, Y., et al., Semicircular canal, saccular and utricular

function in patients with bilateral vestibulopathy: analysis based

on etiology. J Neurol, 2013. 260(3): p. 876-83.

18. McGarvie, L.A., et al., What does the head impulse test versus

caloric dissociation reveal about vestibular dysfunction in

Meniere's disease? Ann N Y Acad Sci, 2015.

19. Guinand, N., et al., Visual acuity while walking and oscillopsia

severity in healthy subjects and patients with unilateral and

bilateral vestibular function loss. Arch Otolaryngol Head Neck

Surg, 2012. 138(3): p. 301-6.

20. Aaronson, N.K., et al., Translation, validation, and norming of the

Dutch language version of the SF-36 Health Survey in community

and chronic disease populations. J Clin Epidemiol, 1998. 51(11):

p. 1055-68.

21. Jacobson, G.P. and C.W. Newman, The development of the

Dizziness Handicap Inventory. Arch Otolaryngol Head Neck Surg,

1990. 116(4): p. 424-7.

22. Tian, J.R., I. Shubayev, and J.L. Demer, Dynamic visual acuity

during passive and self-generated transient head rotation in

normal and unilaterally vestibulopathic humans. Exp Brain Res,

2002. 142(4): p. 486-95.

23. Vital, D., et al., A new dynamic visual acuity test to assess

peripheral vestibular function. Arch Otolaryngol Head Neck Surg,

2010. 136(7): p. 686-91.

24. Fife, T.D., et al., Assessment: vestibular testing techniques in

adults and children: report of the Therapeutics and Technology

Assessment Subcommittee of the American Academy of Neurology.

Neurology, 2000. 55(10): p. 1431-41.

25. Hain, T.C., M. Cherchi, and D.A. Yacovino, Bilateral vestibular

loss. Semin Neurol, 2013. 33(3): p. 195-203.

14

26. Herdman, S.J., et al., Computerized dynamic visual acuity test in

the assessment of vestibular deficits. Am J Otol, 1998. 19(6):

p. 790-6.

27. Kim, S., et al., Bilateral vestibulopathy: clinical characteristics

and diagnostic criteria. Otol Neurotol, 2011. 32(5): p. 812-7.