Development of a new software and test setup for analyzing hVOR in very young children by vHIT

Abstract

INTRODUCTION: Earlier work revealed that vHIT examination is often difficult to perform on very young children. In particular, the calibration of the system can be difficult, as active cooperation of the patient is required. Additionally, the patient must be able to follow the examiner’s instructions, which is challenging for very young children. Therefore, the aim of the present study was to develop and validate a new, software-based approach enabling vHIT testing of young children and infants.

METHODS AND MATERIALS: Six patients (3 boys and 3 girls) aged 5–36 months were included in a prospective, monocentric study between January 2015 and August 2015.

The newly developed intuitive software enabled calibration of the eye position signal with the subjects fixating on animated animal graphics which were projected on a screen. Testing ten healthy adults validated this new calibration and measurement method. After calibration, a vHIT goggle (EyeSeeCam^©) was used to perform head impulses in the horizontal plane while the patient was watching a movie sitting on their parent’s lap or in a baby chair. At least 15 impulses to each side were obtained and the occurrence of refixation saccades was analyzed. All tests were performed by one of two experienced examiners.

RESULTS: The new calibration method and modified test setup provided reproducible results for all patients tested. An increased incidence of artifacts was not observed. In 2 patients, more than one test was needed. None of the included children showed catch-up overt or catch-up covert saccades. There was no gain reduction of more than two standard deviations as compared to the normative results published in the literature on vHIT examinations of children.

CONCLUSION: The proposed protocol allows vHIT testing in very young children and infants (aged 5 months to 3 years). The study emphasizes that vHIT is an easy and sensitive screening tool to evaluate vestibular function in children and should be used as the gold standard in pediatric vestibular assessment.

Keywords

vHIT hVOR dizziness infant vestibular

1 Introduction

The video head impulse test (vHIT) is a well-established procedure to evaluate vestibular function in adults [1, 2]. The test assesses the gain of the angular vestibular-ocular reflex (aVOR) during head rotations as a ratio of eye-to-head acceleration. It is an easy and quick adjunct to evaluate semicircular function. Furthermore, it objectifies and quantifies the clinical head impulse test, first described in 1988 by Halmagy and Curthoys [3].

It has been shown that the incidence of vestibular disorders and imbalance in children and infants is increasing [4 –6]. Since traditional vestibular tests such as rotatory chair testing and caloric tests are usually not tolerated well by children [7], a new, standardized diagnostic protocol for vestibular assessment of children is urgently needed.

A recent study demonstrated that the vHIT is tolerated well in children aged 3–16 years [8]. Reproducible test results were achieved with more than 76% of the participants; the vHIT is an easy-to-use screening tool to evaluate vestibular function not only for adults but also in the pediatric population [8].

Nevertheless, vHIT examination is very difficult to use with very young children. The calibration with laser dots or animal pictures requires the patient’s active cooperation. Additionally, the participant must be able to understand the examiner’s instructions. Arbitrary fixation of a target is crucial to accurate testing and calibration. These requirements are not available with infants (children under 3 years of age). Therefore, the aim of this study was to further improve calibration and the process of vHIT testing of infants younger than 3 years old.

2 Material and methods

Written, informed consent was obtained from those with parental responsibility of all the tested children. The study was conducted in conformity with the Declaration of Helsinki and approved by the ethics committee of the University Medicine Mannheim, Germany.

2.1 Patients and study design

Six children (3 boys and 3 girls) aged 5–36 months were included in this prospective, monocentric study. They were tested between January and August 2015. The children were referred to our pedaudiological department by their pediatricians to further clarify a possible hearing disorder. None of them had been seen by a specialist or had had objective hearing tests before. In our pedaudiologic assessment, all the children showed normal hearing. Likewise, none of the children had been previously diagnosed with a vestibular disorder or a syndromal disease. Each child received a full ENT examination prior to vHITtesting.

Afterwards, the vHIT procedure was explained to the parents. The parents were given swimming goggles for their children to let the children become accustomed to and feel comfortable with wearing goggles. The goggles were simple children’s swimming goggles with clear lenses, obtained at a sports store and provided by the clinic. The parents carried out the required training independently at home for about 2 weeks (1–4 weeks) before they returned to the clinic for further testing.

As a result of this at-home training, no additional time for habituation to the procedure was required during the actual test. For vHIT testing, vHIT goggles (EyeSeeCam^©, Interacoustics, Denmark) were used. The horizontal vestibuloocular reflex was analyzed using vHIT. All tests were performed by 1 of 2 experienced examiners.

2.2 Software design

The vHIT examination is exceedingly difficult to conduct on very young children. Children must feel comfortable to sustain the positions necessary to take accurate readings. Calibration requires the patient’s active cooperation. To achieve this, a calibration and measurement software which runs on standard laptop hardware was developed. During vHIT, a laptop running the software was connected to a large screen in front of the patient. To avoid distraction, the software ran in full screen mode so that the operating system was hidden and only relevant content was visible. The software was controlled using keyboard commands intuitive for the examiner and less distracting to children than clickable control elements (buttons) in the visual interface. The object-oriented software design is easily maintained and open for extension. The steps of the vHIT examination (comforting, calibration, measuring, and pauses) are encapsulated in corresponding software states sharing a common interface following the well-known state software design pattern such that, for example, integration of new examination steps for specific clinical questions is straightforward and does not require error-prone modification of the existing states. The sequence of states in the software application and the state transitions are illustrated in an activity diagram in Table 1. C++ and the cross-platform application framework Qt were used to implement the software, which was successfully tested on Windows 7, Windows 8, and Linux.

Table 1
Activity diagram of the developed software

2.3 Calibration

Calibration was necessary to ensure that head and eye movements overlaid. Prior to calibration, vHIT goggle slippage was minimized as much as possible; the vHIT goggles were tightened on the head until movement of the goggles at the bridge of the nose and the lateral zygomatic bone was at an absolute minimum. Calibration of the eye position signal was carried out with the subjects successively fixating on animated animal graphics projected on the screen and separated by a known horizontal angle. During calibration, the head was fixed. The room was dimly lit to ensure the subjects’ interest in and visual focus on the icons. Low, predictable pupil velocities during head fixation ensured valid calibration on the x and y axes. The animal graphics appeared in the central, left, right, and bottom zones of the screen at a given distance and angle.

2.4 Test-position

Compared to the standard vHIT procedure, the test setup was more aligned to the patient. The measurement screen was adjustable in the room. This allows for measuring of bedridden patients and smallchildren in infant carriers, among others. Traditionally, examination of infants under the age of 1 year has not been possible, as most children learn to sit freely between the 6 and 9 months of age and the traditional test setup is for patients sitting upright. Furthermore, older children who can sit on their own are often intimidated in unfamiliar situations and prefer not to sit alone in a chair. Hence, for children who cannot sit on their own, the current study employed vHIT testing using a commercial baby seat (Mico AP, Dorel TM, Canada). A baby seat allows for sufficient head rotation with the best possible spinal fixation for infants from 5 to 22 pounds and up to 29 inches. It allows parents to sit next to the children and give them the confidence to sit through the examination. Children old enough to sit on their own but too intimidated to do so for the examination were examined in a special sitting position: the parent sat in a chair with the patient sitting sideways in his or her lap, with the patient’s legs fixed between the parent’s legs. The patient placed one arm around the parent’s back and the other fixed on the parent’s abdomen. The close contact of the child’s upper body with the parent’s torso made for an enjoyable position for both child and parent, and the position provided good accessibility to the patient’s head for the examiner (Fig. 1).

Fig.1

This figure demonstrates the test set-up and shows testing of a 3 year old girl sitting on her mother’s lap (above) and testing of a 5 and 10 months old child (below).

2.5 Validation

Ten adults without any vestibular disorder were tested, validating the new calibration and test setup. The results of the new method were compared to the results of a laser dot–based, standard calibration system which is integrated in the EyeSeeCam^© system.Each method was performed 3 times for each of the 10 adults (60 tests in total). The standard calibration and measurement method showed normally distributed gains averaging 0.94 with a standard deviation of 0.07. The results of the vHIT examination using the newly developed calibration and measurement software showed normally distributed gains averaging 0.93 with standard deviation of 0.08. This corresponds to an intraclass correlation coefficient of 0.96 and a Pearson coefficient of 0.92. The relatively large values of the coefficients demonstrate a high congruence of the calibration results using the standard calibration system and the newly developed calibration software (Fig. 2).

Fig.2

This figure shows one of the validated calibrations.

2.6 vHIT

Horizontal VOR was measured using a portable, lightweight vHIT goggle device (EyeSeeCam^©) system (Interacoustics A/S) comprising a high-speed, infrared camera (sampling rate of 250 Hz) and a built-in accelerometer. vHIT was performed with the child in a seated position in a dimly lit room. Calibration was performed with the new software program. After successful calibration, the child focused on a far target (distance approximately 1.5 m) depicting an animal graphic. Target head velocity was 100 to 200 degrees per second with amplitude between 5 and 15 degrees from center to lateral. The impulses were unpredictable with respect to both directions (right and left). The patient was given a break as needed after an impulse to resume concentration. In total, at least 15 impulses to each side were obtained. VOR gain was calculated as the ratio of eye to head velocity at 40, 60, and 80 msec. Gain variance was assessed and the appearance of covert and overt saccades were analyzed. A compensatory saccade was classified as a “covert” saccade when it occurred during the head movement [9]. If the saccade occurred after the head velocity crossed zero, it was considered “overt.”

A paired, one-sided t-test was performed to analyze the differences in gain between 40, 60, and 80 msec. Descriptive statistics were used to measure VOR gains, time of examination, and gain variance. For all statistical analyses, associations were considered statistically significant with a value of p < 0.05. Statistical analysis was performed using SPSS Statistics 17.0 (SPSS Inc., Chicago, IL USA).

3 Results

Meaningful and reproducible results for all performed measurements were obtained with the presented calibration method and test setup. An increased incidence of artifacts was not observed. In 2 patients, more than 1 test was needed. Two subjects were examined in the supine position in the child seat. vHIT of the remaining 4 children was performed in the sitting position in the parent’s lap as described above. There were no differences between the 2 examination positions with respect to vHIT calibration, execution, or measurement results. The study time required was extended about 5 minutes for the new calibration and test setup as compared to adult examinations, but explanation and patient cooperation were not necessary. The mean time expenditure was approximately 13 minutes, which is significantly shorter than that of the classical testing of children including calibration and vHIT testing with animal pictograms (approximately 20 minutes), and significantly longer than that of cognitive healthy adults (approximately 8 minutes).

None of the included subjects showed catch-up overt or catch-up covert saccades. There was no gain reduction of more than 2 standard deviations as compared to the published normative results for child subjects.

The median gain was 0.78 (+/–0.23). The median gain for both sides was 1.06 +/–0.25 after 40 msec, 0.76 +/–0.21 after 60 msec, and 0.53 +/–0.21 after 80 msec. The detected gain for impulses to the right was 1.06 +/–0.22 after 40 msec, 0.74 +/–0.2 after 60 msec, and 0.57 +/–0.18 after 80 msec. The detected gain for impulses to the left was 1.05 +/, 0.29 after 40 msec, 0.79 +/–0.21 after 60 msec, and 0.53 +/–0.24 after 80 msec (Fig. 3).

Fig.3

(a) 30 months old patient measured and calibrated with the standard system. (b) Results of vHIT testing of the same patient 2 months later with the presented new system.

4 Discussion

The incidence of vestibular disorders in children has been on the rise [5, 10]. Given that a loss of vestibular function can affect the entire development of a child [11, 12] and may lead to severe problems in school and in motor development, a valid and reliable tool for vestibular assessment in children is urgently needed to identify children with vestibular impairment. Since vHIT has been successfully performed in children aged between 3 and 16 years [8], the current study sought to further improve vHIT-testing in very young children and discover an advanced method to diagnose vestibular hypofunction in infants. If diagnosed early, vestibular physical therapy can stimulate the compensation mechanism of the visual and proprioceptive sensory system [11]. Therefore, early diagnosis should be a target of pediatricians and otorhinolaryngologists.

Vestibular assessment plays an important role in many preventative and pre- and post-operative diagnoses. It should be part of routinely-performed diagnostics prior to cochlear implantation in children with congenital or acquired deafness. Hülse et al. state that only accurate, pre-operative vestibular testing allows evaluation of potential post-operative dizziness after cochlea implantation and should be documented precisely [12]. Vestibular assessment is also crucial to identify possible vestibular prosthesis implant patients who would benefit from a simultaneous implantation of a cochlea implant and vestibular prosthesis.

Alternatives to vHIT are inadequate to address young patients’ needs. Vestibular evoked myogenic potentials (VEMPs) and caloric and rotatory tests are sufficient to evaluate vestibular function in children [13, 14], but clinical experience shows that caloric and rotary tests in particular are not tolerated well in children and take much longer than vHIT.

Earlier studies on testing children concluded that the calibration of the vHIT goggles is the mostsusceptible part of the test. Children under 6 years of age in particular have difficulty following an examiner’s instructions to look to the right, left, center, up, or down and to focus on laser dots. Additionally, some infants are unable to maintain a stable head position during calibration; consequently, the laser dots used for calibration would move and blur in their vision. Furthermore, many young patients are not able to distinguish right from left in order to follow directional instructions but are able to fixate on different animal icons quite easily [8].

This finding led to the present study’s protocol for calibration. The presented software and test setup offers an objective and validated method for testing the vestibular organ in very young children. The intuitive design of the presented method stimulates the child’s curiosity and allows an investigation of infants. The design is also useful in the study of deaf children or children who do not have adequate language comprehension or vocabulary to follow oral instruction.

Furthermore, the modified vHIT might also be helpful in in frightened or traumatized children, such as is found in the increasing population of children forced to flee from war or war-like conditions. Verbal communication with these patients is very difficult, but the presented method offers a working alternative. Such methods of investigation appear to be particularly important in these patient populations because communication disorders and intellectual disability are often later or non-recognized [15]. In the few publications examining physical trauma, the most likely causes of injuries in children are war against children, bombing, and landmines [16, 17]. Bombing and landmines can also lead to vestibularcochlear injuries. Because of this, vestibular testing in refugee children is necessary.

The new calibration protocol renders the test easier to accomplish, as only 1 examiner is required to compete it. The infant should be sitting on the lap of a family member in front of the screen, which should be at a distance of approximately 1.5 m. After the goggles are placed correctly, the calibration can begin and testing can proceed. With the new protocol, calibration had to be repeated in two patients. The remaining infants (N = 4) could be easily tested after one successful trial of calibration. Therefore vHIT testing took approximately 20 minutes in total, which underlines how quick the method is. The vestibular system is anatomically developed and functioning at birth, though vestibular responses may be variable in the newborn [13 , 19]. VOR responses in the neonatal infant are poorly developed 24 to 120 hours postnatal, but will usually normalize within the first two months [20]. From that point on, vestibular function may be assessed. However, for assessing vHIT the subject must be able to focus on a nearby target [21, 22]. The ability to focus on a steady point is the limiting factor for vHIT assessment of infants. If necessary, several testing trials with the new calibration program can be performed to accustom the patient to thetask.

Median gain in our tested children was 0.78, approximately 2 standard deviations lower than the gain measured in adults. Covert and overt saccades were not detected in the tested infants. As saccades can be found in older individuals without any sign of vestibular loss [23], they seem to be pathologic in children of every age [8]. It is therefore concluded that vHIT testing is a good screening tool to accurately detect vestibular hypofunction in children. Further assessment including VEMP testing or subjective visual vertical (SVV) test and also a hearing assessment should be performed if a vestibular disorder is suspected.

Overall, vHIT appears to be the most efficient and sensitive screening method to quantify vestibular function, regardless of age. Therefore, future studies should try to detect the subtle differences in vHIT results among age groups to learn more about changes in VOR caused by aging. The present study is limited to a very small sample size but the new calibration program shows that vestibular testing is doable in very young children. Further work should focus on both ends of the age range, as the test results of those two groups deviate the most from the norm.

5 Conclusion

vHIT is an easy and sensitive screening tool to evaluate vestibular function in children and should be used as the gold standard in pediatric vestibular assessment. The proposed protocol improves and simplifies vHIT testing in very young children and infants (aged 5 months to 3 years).

References

Weber

K.P.

, MacDougall

H.G.

, Halmagyi

G.M.

and Curthoys

I.S.

, Impulsive testing of semicircular-canal function using video-oculography, Ann N Y Acad Sci 1164 (2009), 486–491.

Walther

P.D.L.

, Hörmann

, Bloching

and Blödow

, Rezeptorfunktion der Bogengänge, HNO 60(1) (2012), 75–88.

Halmagyi

G.M.

and Curthoys

I.S.

, A clinical sign of canal paresis, Arch Neurol 45(7) (1988), 737–739.

Zannolli

, Zazzi

, Muraca

M.C.

, Macucci

, Buoni

and Nuti

, A child with vestibular neuritis. Is adenovirus implicated?, Brain and Development 28(6) (2006), 410–412.

Rine

R.M.

, Growing evidence for balance and vestibular problems in children, Audiological Medicine 7(3) (2009), 138–142.

Saka

, Imai

, Seo

, Ohta

, Fujimori

, Masumura

, Inohara

and Sakagami

, Analysis of benign paroxysmal positional nystagmus in children, International Journal of Pediatric Otorhinolaryngology 77(2) (2013), 233–236.

Gioacchini

F.M.

, Alicandri-Ciufelli

, Kaleci

, Magliulo

and Re

, Prevalence and diagnosis of vestibular disorders in children: A review, International Journal of Pediatric Otorhinolaryngology 78(5) (2014), 718–724.

Hulse

, Hormann

, Servais

J.J.

, Hulse

and Wenzel

, Clinical experience with video Head Impulse Test in children, Int J Pediatr Otorhinolaryngol 79(8) (2015), 1288–1293.

Weber

, Aw

, Todd

, McGarvie

, Curthoys

and Halmagyi

, Head impulse test in unilateral vestibular loss Vestibulo-ocular reflex and catch-up saccades, Neurology 70(6) (2008), 454–463.

10.

Wiener-Vacher

S.R.

, Vestibular disorders in children, Int J Audiol 47(9) (2008), 578–583.

11.

Kaga

, Rehabilitation of Children with Disequilibrium. Vertigo and Balance Disorders in Children, Springer, (2014)63–68.

12.

Hülse

, Hülse

, Wenzel

, Hörmann

and Servais

, Frequency-specific Analysis of the hVOR Prior of Cochlear Implant Operation, Laryngo-Rhino-Otologie 94(3) (2015), 173–178.

13.

Eviatar

and Eviatar

, The normal nystagmic response of infants to caloric and perrotatory stimulation, Laryngoscope 89(7 Pt 1) (1979), 1036–1045.

14.

Kelsch

T.A.

, Schaefer

L.A.

and Esquivel

C.R.

, Vestibular evoked myogenic potentials in young children: Test parameters and normative data, Laryngoscope 116(6) (2006), 895–900.

15.

Miles

, Afghan children and mental retardation: Information, advocacy and prospects, Disabil Rehabil 19(11) (1997), 496–500.

16.

Pearn

, Children and war, J Paediatr Child Health 39(3) (2003), 166–172.

17.

Celikel

, Karbeyaz

, Kararslan

, Arslan

M.M.

and Zeren

, Childhood casualties during civil war: Syrian experience, J Forensic Leg Med 34 (2015), 1–4.

18.

Mulch

and Petermann

, Influence of age on results of vestibular function tests. Review of literature and presentation of caloric test results, Ann Otol Rhinol Laryngol Suppl 88(2 Pt 2 Suppl 56) (1979), 1–17.

19.

Ornitz

E.M.

, Atwell

C.W.

, Walter

D.O.

, Hartmann

E.E.

and Kaplan

A.R.

, The maturation of vestibular nystagmus in infancy and childhood, Acta Otolaryngol 88(3-4) (1979), 244–256.

20.

Weissman

B.M.

, DiScenna

A.O.

and Leigh

R.J.

, Maturation of the vestibulo-ocular reflex in normal infants during the first 2 months of life, Neurology 39(4) (1989), 534–538.

21.

Aslin

R.N.

, Development of binocular fixation in human infants, Journal of Experimental Child Psychology 23(1) (1977), 133–150.

22.

Thorn

, Gwiazda

, Cruz

, Bauer

J.A.

and Held

, The development of eye alignment, convergence, and sensory binocularity in young infants, Invest Ophthalmol Vis Sci 35(2) (1994), 544–553.

23.

Agrawal

, Schubert

M.C.

, Migliaccio

A.A.

, Zee

D.S.

, Schneider

, Lehnen

and Carey

J.P.

, Evaluation of quantitative head impulse testing using search coils versus video-oculography in older individuals, Otology & Neurotology 35(2) (2014), 283–288.